ベイズ最適化および ASHA 最適化による分類器の自動選択
この例では、fitcauto を使用し、指定した学習予測子と応答データに基づいてさまざまなハイパーパラメーターの値をもつ分類モデルのタイプの選択を自動的に試す方法を示します。既定では、この関数はモデルの選択と評価にベイズ最適化を使用します。学習データ セットに多数の観測値が含まれている場合は、代わりに非同期連続半減アルゴリズム (ASHA) を使用できます。最適化が完了すると、fitcauto は、データ セット全体で学習済みの、新しいデータの分類に最適であると予測したモデルを返します。テスト データに対するモデルの性能をチェックします。
標本データの読み込み
この例では census1994.mat に保存されている 1994 年の国勢調査データを使用します。このデータ セットは、個人の年収が $50,000 を超えるかどうかを予測するために使用できる、米国国勢調査局の人口統計情報から構成されています。
学習データ adultdata およびテスト データ adulttest を含む、標本データ census1994 を読み込みます。学習データ セットの最初の数行をプレビューします。
load census1994
head(adultdata) age workClass fnlwgt education education_num marital_status occupation relationship race sex capital_gain capital_loss hours_per_week native_country salary
___ ________________ __________ _________ _____________ _____________________ _________________ _____________ _____ ______ ____________ ____________ ______________ ______________ ______
39 State-gov 77516 Bachelors 13 Never-married Adm-clerical Not-in-family White Male 2174 0 40 United-States <=50K
50 Self-emp-not-inc 83311 Bachelors 13 Married-civ-spouse Exec-managerial Husband White Male 0 0 13 United-States <=50K
38 Private 2.1565e+05 HS-grad 9 Divorced Handlers-cleaners Not-in-family White Male 0 0 40 United-States <=50K
53 Private 2.3472e+05 11th 7 Married-civ-spouse Handlers-cleaners Husband Black Male 0 0 40 United-States <=50K
28 Private 3.3841e+05 Bachelors 13 Married-civ-spouse Prof-specialty Wife Black Female 0 0 40 Cuba <=50K
37 Private 2.8458e+05 Masters 14 Married-civ-spouse Exec-managerial Wife White Female 0 0 40 United-States <=50K
49 Private 1.6019e+05 9th 5 Married-spouse-absent Other-service Not-in-family Black Female 0 0 16 Jamaica <=50K
52 Self-emp-not-inc 2.0964e+05 HS-grad 9 Married-civ-spouse Exec-managerial Husband White Male 0 0 45 United-States >50K
各行には、成人 1 人の人口統計情報が格納されています。最後の列 salary は個人の年収が $50,000 以下か、$50,000 を超えるかどうかを示します。
adultdata と adulttest から欠損値を含む観測値を削除します。
adultdata = rmmissing(adultdata); adulttest = rmmissing(adulttest);
ベイズ最適化による自動モデル選択の使用
fitcauto を使用して、adultdata のデータに適切な分類器を見つけます。既定では、fitcauto はベイズ最適化を使用してモデルとそのハイパーパラメーターの値を選択し、各モデルの交差検証の分類誤差 (Validation loss) を計算します。既定の設定では、fitcauto は、最適化のプロット、および最適化の結果の反復表示を提供します。これらの結果を解釈する方法の詳細については、Verbose の表示を参照してください。
観測値の重みを設定し、ベイズ最適化を並列実行するよう指定します。これには Parallel Computing Toolbox™ が必要です。並列でのタイミングに再現性がないため、並列ベイズ最適化で再現性のある結果が生成されるとは限りません。最適化の複雑度に応じて、特に大きなデータ セットでは、この処理に時間がかかる場合があります。
bayesianOptions = struct("UseParallel",true); [bayesianMdl,bayesianResults] = fitcauto(adultdata,"salary","Weights","fnlwgt", ... "HyperparameterOptimizationOptions",bayesianOptions);
Warning: Data set has more than 10000 observations. Because ASHA optimization often finds good solutions faster than Bayesian optimization for data sets with many observations, try specifying the 'Optimizer' field value as 'asha' in the 'HyperparameterOptimizationOptions' value structure.
Starting parallel pool (parpool) using the 'Processes' profile ... Connected to parallel pool with 6 workers. Copying objective function to workers... Done copying objective function to workers. Learner types to explore: ensemble, nb, svm, tree Total iterations (MaxObjectiveEvaluations): 120 Total time (MaxTime): Inf |=======================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Estimated min | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | validation loss | | | |=======================================================================================================================================================| | 1 | 6 | Best | 0.24677 | 43.754 | 0.24677 | 0.24677 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 207 | | | | | | | | | | MinLeafSize: 8946 | | 2 | 6 | Accept | 0.24677 | 0.88853 | 0.24677 | 0.24677 | tree | MinLeafSize: 6845 | | 3 | 6 | Best | 0.14543 | 2.2401 | 0.14543 | 0.19002 | tree | MinLeafSize: 45 | | 4 | 6 | Accept | 0.15419 | 80.045 | 0.14543 | 0.19002 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 209 | | | | | | | | | | MinLeafSize: 2 | | 5 | 6 | Accept | 0.14928 | 93.473 | 0.14543 | 0.18783 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 201 | | | | | | | | | | MinLeafSize: 480 | | 6 | 6 | Accept | 0.15158 | 44.195 | 0.14543 | 0.15158 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.0088611 | | | | | | | | | | Standardize: false | | 7 | 6 | Accept | 0.14889 | 139.61 | 0.14543 | 0.15158 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 297 | | | | | | | | | | MinLeafSize: 510 | | 8 | 6 | Accept | 0.18439 | 1.4145 | 0.14543 | 0.15158 | tree | MinLeafSize: 3104 | | 9 | 6 | Accept | 0.14554 | 161.4 | 0.14543 | 0.14554 | svm | BoxConstraint: 0.65272 | | | | | | | | | | KernelScale: 6.0716 | | 10 | 6 | Accept | 0.15127 | 166.83 | 0.14543 | 0.14881 | svm | BoxConstraint: 0.29711 | | | | | | | | | | KernelScale: 12.637 | | 11 | 6 | Accept | 0.16715 | 76.475 | 0.14543 | 0.14881 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.22835 | | | | | | | | | | Standardize: true | | 12 | 6 | Accept | 0.20124 | 101.74 | 0.14543 | 0.14881 | nb | DistributionNames: kernel | | | | | | | | | | Width: 18.24 | | | | | | | | | | Standardize: true | | 13 | 6 | Accept | 0.24677 | 0.46731 | 0.14543 | 0.14881 | tree | MinLeafSize: 12859 | | 14 | 6 | Accept | 0.168 | 0.69093 | 0.14543 | 0.14881 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 15 | 6 | Accept | 0.24677 | 3.7861 | 0.14543 | 0.17365 | svm | BoxConstraint: 0.24765 | | | | | | | | | | KernelScale: 0.0016358 | | 16 | 6 | Accept | 0.14757 | 1.8122 | 0.14543 | 0.17365 | tree | MinLeafSize: 135 | | 17 | 6 | Accept | 0.168 | 0.36422 | 0.14543 | 0.17099 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 18 | 6 | Accept | 0.23488 | 55.736 | 0.14543 | 0.17099 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 243 | | | | | | | | | | MinLeafSize: 5284 | | 19 | 6 | Accept | 0.24677 | 48.557 | 0.14543 | 0.17099 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 217 | | | | | | | | | | MinLeafSize: 6309 | | 20 | 6 | Accept | 0.16312 | 44.074 | 0.14543 | 0.16836 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.043878 | | | | | | | | | | Standardize: true | |=======================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Estimated min | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | validation loss | | | |=======================================================================================================================================================| | 21 | 6 | Accept | 0.19857 | 99.234 | 0.14543 | 0.17104 | nb | DistributionNames: kernel | | | | | | | | | | Width: 3.1517 | | | | | | | | | | Standardize: true | | 22 | 6 | Accept | 0.15361 | 1.5262 | 0.14543 | 0.16123 | tree | MinLeafSize: 338 | | 23 | 6 | Accept | 0.15417 | 92.662 | 0.14543 | 0.16123 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 277 | | | | | | | | | | MinLeafSize: 101 | | 24 | 6 | Accept | 0.1628 | 2.9924 | 0.14543 | 0.15776 | tree | MinLeafSize: 6 | | 25 | 6 | Accept | 0.15479 | 85.937 | 0.14543 | 0.15776 | nb | DistributionNames: kernel | | | | | | | | | | Width: 4.0678 | | | | | | | | | | Standardize: false | | 26 | 6 | Best | 0.1445 | 1.7904 | 0.1445 | 0.15547 | tree | MinLeafSize: 105 | | 27 | 6 | Accept | 0.15763 | 2.8146 | 0.1445 | 0.15223 | tree | MinLeafSize: 9 | | 28 | 6 | Accept | 0.24677 | 1.9596 | 0.1445 | 0.15223 | svm | BoxConstraint: 0.0047884 | | | | | | | | | | KernelScale: 0.0048876 | | 29 | 6 | Accept | 0.15158 | 42.575 | 0.1445 | 0.15223 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.0075341 | | | | | | | | | | Standardize: false | | 30 | 6 | Accept | 0.16813 | 1.1328 | 0.1445 | 0.15262 | tree | MinLeafSize: 598 | | 31 | 6 | Accept | 0.14994 | 2.2018 | 0.1445 | 0.15084 | tree | MinLeafSize: 21 | | 32 | 6 | Accept | 0.168 | 0.4177 | 0.1445 | 0.15084 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 33 | 6 | Accept | 0.14535 | 165.79 | 0.1445 | 0.15084 | svm | BoxConstraint: 0.14315 | | | | | | | | | | KernelScale: 3.2371 | | 34 | 6 | Accept | 0.17901 | 5.6933 | 0.1445 | 0.1502 | tree | MinLeafSize: 2 | | 35 | 6 | Accept | 0.168 | 0.48234 | 0.1445 | 0.1502 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 36 | 6 | Accept | 0.16795 | 1.0938 | 0.1445 | 0.15088 | tree | MinLeafSize: 570 | | 37 | 6 | Accept | 0.18403 | 0.65343 | 0.1445 | 0.15009 | tree | MinLeafSize: 2119 | | 38 | 6 | Accept | 0.24677 | 2.0224 | 0.1445 | 0.15009 | svm | BoxConstraint: 216.63 | | | | | | | | | | KernelScale: 0.006762 | | 39 | 6 | Accept | 0.14945 | 83.448 | 0.1445 | 0.15009 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 223 | | | | | | | | | | MinLeafSize: 345 | | 40 | 6 | Accept | 0.16129 | 90.614 | 0.1445 | 0.15009 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 278 | | | | | | | | | | MinLeafSize: 388 | |=======================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Estimated min | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | validation loss | | | |=======================================================================================================================================================| | 41 | 6 | Accept | 0.1647 | 32.478 | 0.1445 | 0.15009 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 256 | | | | | | | | | | MinLeafSize: 7032 | | 42 | 6 | Accept | 0.16698 | 71.592 | 0.1445 | 0.15009 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 226 | | | | | | | | | | MinLeafSize: 861 | | 43 | 6 | Accept | 0.15872 | 189.88 | 0.1445 | 0.15009 | svm | BoxConstraint: 11.079 | | | | | | | | | | KernelScale: 155.11 | | 44 | 6 | Best | 0.14353 | 1.687 | 0.14353 | 0.14816 | tree | MinLeafSize: 79 | | 45 | 6 | Accept | 0.15374 | 39.891 | 0.14353 | 0.14816 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.0035642 | | | | | | | | | | Standardize: true | | 46 | 6 | Accept | 0.14979 | 83.79 | 0.14353 | 0.14816 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 203 | | | | | | | | | | MinLeafSize: 911 | | 47 | 6 | Accept | 0.24677 | 2.0222 | 0.14353 | 0.14816 | svm | BoxConstraint: 2.0475 | | | | | | | | | | KernelScale: 0.0049311 | | 48 | 6 | Accept | 0.20139 | 94.777 | 0.14353 | 0.14816 | nb | DistributionNames: kernel | | | | | | | | | | Width: 93.844 | | | | | | | | | | Standardize: true | | 49 | 6 | Accept | 0.168 | 0.38418 | 0.14353 | 0.14816 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 50 | 6 | Accept | 0.15146 | 47.829 | 0.14353 | 0.14816 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.36089 | | | | | | | | | | Standardize: false | | 51 | 6 | Accept | 0.1686 | 27.403 | 0.14353 | 0.14816 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 261 | | | | | | | | | | MinLeafSize: 8358 | | 52 | 6 | Accept | 0.23867 | 195.4 | 0.14353 | 0.14816 | svm | BoxConstraint: 0.011694 | | | | | | | | | | KernelScale: 18.422 | | 53 | 6 | Accept | 0.15319 | 162.54 | 0.14353 | 0.14816 | svm | BoxConstraint: 0.010059 | | | | | | | | | | KernelScale: 4.0044 | | 54 | 6 | Accept | 0.15501 | 79.775 | 0.14353 | 0.14816 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 232 | | | | | | | | | | MinLeafSize: 24 | | 55 | 6 | Accept | 0.16874 | 2864.2 | 0.14353 | 0.14816 | svm | BoxConstraint: 2.3437 | | | | | | | | | | KernelScale: 1.5289 | | 56 | 6 | Accept | 0.15742 | 2870.4 | 0.14353 | 0.14816 | svm | BoxConstraint: 586.48 | | | | | | | | | | KernelScale: 6.2332 | | 57 | 6 | Accept | 0.15119 | 81.33 | 0.14353 | 0.14816 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 209 | | | | | | | | | | MinLeafSize: 1622 | | 58 | 6 | Accept | 0.16761 | 3059.5 | 0.14353 | 0.14816 | svm | BoxConstraint: 380.05 | | | | | | | | | | KernelScale: 3.8304 | | 59 | 6 | Accept | 0.1522 | 77.374 | 0.14353 | 0.14816 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 204 | | | | | | | | | | MinLeafSize: 1676 | | 60 | 6 | Accept | 0.17435 | 3048.3 | 0.14353 | 0.14816 | svm | BoxConstraint: 1.424 | | | | | | | | | | KernelScale: 1.1302 | |=======================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Estimated min | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | validation loss | | | |=======================================================================================================================================================| | 61 | 6 | Accept | 0.15944 | 62.648 | 0.14353 | 0.14816 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 228 | | | | | | | | | | MinLeafSize: 2945 | | 62 | 6 | Accept | 0.1622 | 2716.5 | 0.14353 | 0.14816 | svm | BoxConstraint: 8.1184 | | | | | | | | | | KernelScale: 2.3874 | | 63 | 6 | Accept | 0.15986 | 53.967 | 0.14353 | 0.14816 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 203 | | | | | | | | | | MinLeafSize: 3145 | | 64 | 6 | Accept | 0.14553 | 150.62 | 0.14353 | 0.14816 | svm | BoxConstraint: 0.26261 | | | | | | | | | | KernelScale: 5.0699 | | 65 | 6 | Accept | 0.14799 | 157.27 | 0.14353 | 0.14816 | svm | BoxConstraint: 0.081785 | | | | | | | | | | KernelScale: 4.9516 | | 66 | 6 | Accept | 0.14833 | 158.24 | 0.14353 | 0.14816 | svm | BoxConstraint: 0.74137 | | | | | | | | | | KernelScale: 12.172 | | 67 | 6 | Accept | 0.20535 | 205.52 | 0.14353 | 0.14816 | svm | BoxConstraint: 0.26823 | | | | | | | | | | KernelScale: 60.591 | | 68 | 6 | Accept | 0.15533 | 173.46 | 0.14353 | 0.14816 | svm | BoxConstraint: 0.031445 | | | | | | | | | | KernelScale: 7.4327 | | 69 | 6 | Accept | 0.16052 | 179.49 | 0.14353 | 0.14816 | svm | BoxConstraint: 0.0062986 | | | | | | | | | | KernelScale: 4.855 | | 70 | 6 | Accept | 0.16085 | 46.246 | 0.14353 | 0.14816 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 211 | | | | | | | | | | MinLeafSize: 3991 | | 71 | 6 | Accept | 0.15032 | 159.22 | 0.14353 | 0.14816 | svm | BoxConstraint: 0.34417 | | | | | | | | | | KernelScale: 12.342 | | 72 | 6 | Accept | 0.18532 | 46.838 | 0.14353 | 0.14816 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 202 | | | | | | | | | | MinLeafSize: 4515 | | 73 | 6 | Accept | 0.18875 | 189.42 | 0.14353 | 0.14816 | svm | BoxConstraint: 0.0019135 | | | | | | | | | | KernelScale: 5.2624 | | 74 | 6 | Accept | 0.15372 | 161.15 | 0.14353 | 0.14816 | svm | BoxConstraint: 0.2716 | | | | | | | | | | KernelScale: 15.245 | | 75 | 6 | Accept | 0.15399 | 159.07 | 0.14353 | 0.14816 | svm | BoxConstraint: 0.01694 | | | | | | | | | | KernelScale: 4.9823 | | 76 | 6 | Accept | 0.15324 | 160.44 | 0.14353 | 0.1477 | svm | BoxConstraint: 0.35465 | | | | | | | | | | KernelScale: 16.266 | | 77 | 6 | Accept | 0.14811 | 149.74 | 0.14353 | 0.1474 | svm | BoxConstraint: 0.13086 | | | | | | | | | | KernelScale: 5.7377 | | 78 | 6 | Accept | 0.15523 | 165.32 | 0.14353 | 0.14769 | svm | BoxConstraint: 0.18542 | | | | | | | | | | KernelScale: 15.367 | | 79 | 6 | Accept | 0.15441 | 160.44 | 0.14353 | 0.14811 | svm | BoxConstraint: 0.025242 | | | | | | | | | | KernelScale: 5.8741 | | 80 | 6 | Accept | 0.15792 | 169.29 | 0.14353 | 0.14816 | svm | BoxConstraint: 0.013558 | | | | | | | | | | KernelScale: 5.9968 | |=======================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Estimated min | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | validation loss | | | |=======================================================================================================================================================| | 81 | 6 | Accept | 0.14914 | 151.25 | 0.14353 | 0.14816 | svm | BoxConstraint: 1.0545 | | | | | | | | | | KernelScale: 15.459 | | 82 | 6 | Accept | 0.18184 | 48.24 | 0.14353 | 0.14816 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 205 | | | | | | | | | | MinLeafSize: 4097 | | 83 | 6 | Accept | 0.14507 | 164.62 | 0.14353 | 0.14715 | svm | BoxConstraint: 0.29922 | | | | | | | | | | KernelScale: 4.0302 | | 84 | 6 | Accept | 0.15297 | 159.83 | 0.14353 | 0.14669 | svm | BoxConstraint: 0.42207 | | | | | | | | | | KernelScale: 17.016 | | 85 | 6 | Accept | 0.14538 | 164.56 | 0.14353 | 0.14584 | svm | BoxConstraint: 0.30157 | | | | | | | | | | KernelScale: 3.9176 | | 86 | 6 | Accept | 0.15171 | 146.64 | 0.14353 | 0.14681 | svm | BoxConstraint: 0.011941 | | | | | | | | | | KernelScale: 3.7945 | | 87 | 6 | Accept | 0.15038 | 159.26 | 0.14353 | 0.14624 | svm | BoxConstraint: 0.8291 | | | | | | | | | | KernelScale: 17.344 | | 88 | 6 | Accept | 0.14811 | 145.43 | 0.14353 | 0.14638 | svm | BoxConstraint: 2.3513 | | | | | | | | | | KernelScale: 17.792 | | 89 | 6 | Accept | 0.14957 | 145.61 | 0.14353 | 0.14614 | svm | BoxConstraint: 1.2721 | | | | | | | | | | KernelScale: 17.942 | | 90 | 6 | Accept | 0.43333 | 3817 | 0.14353 | 0.14739 | svm | BoxConstraint: 0.46095 | | | | | | | | | | KernelScale: 0.11266 | | 91 | 6 | Accept | 0.14777 | 148.7 | 0.14353 | 0.14651 | svm | BoxConstraint: 3.0283 | | | | | | | | | | KernelScale: 18.144 | | 92 | 6 | Accept | 0.37513 | 2462.6 | 0.14353 | 0.14633 | svm | BoxConstraint: 0.15218 | | | | | | | | | | KernelScale: 0.15693 | | 93 | 6 | Accept | 0.37822 | 2568 | 0.14353 | 0.14674 | svm | BoxConstraint: 0.085986 | | | | | | | | | | KernelScale: 0.15131 | | 94 | 6 | Accept | 0.31891 | 2559 | 0.14353 | 0.14582 | svm | BoxConstraint: 0.142 | | | | | | | | | | KernelScale: 0.14667 | | 95 | 6 | Accept | 0.35037 | 2484.8 | 0.14353 | 0.14719 | svm | BoxConstraint: 732.35 | | | | | | | | | | KernelScale: 0.30475 | | 96 | 6 | Accept | 0.20831 | 3290.1 | 0.14353 | 0.14655 | svm | BoxConstraint: 0.36082 | | | | | | | | | | KernelScale: 0.55221 | | 97 | 6 | Accept | 0.24678 | 4132.6 | 0.14353 | 0.14626 | svm | BoxConstraint: 0.24182 | | | | | | | | | | KernelScale: 0.032229 | | 98 | 6 | Accept | 0.21716 | 3318.5 | 0.14353 | 0.14621 | svm | BoxConstraint: 0.66883 | | | | | | | | | | KernelScale: 0.59271 | | 99 | 6 | Accept | 0.14578 | 248.04 | 0.14353 | 0.14582 | svm | BoxConstraint: 0.4278 | | | | | | | | | | KernelScale: 3.0737 | | 100 | 6 | Accept | 0.15224 | 162.91 | 0.14353 | 0.14686 | svm | BoxConstraint: 1.1624 | | | | | | | | | | KernelScale: 24.402 | |=======================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Estimated min | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | validation loss | | | |=======================================================================================================================================================| | 101 | 6 | Accept | 0.22215 | 3344 | 0.14353 | 0.14576 | svm | BoxConstraint: 0.17981 | | | | | | | | | | KernelScale: 0.41659 | | 102 | 6 | Accept | 0.14999 | 160.35 | 0.14353 | 0.1458 | svm | BoxConstraint: 1.1293 | | | | | | | | | | KernelScale: 18.197 | | 103 | 6 | Accept | 0.14671 | 265.17 | 0.14353 | 0.14599 | svm | BoxConstraint: 0.29559 | | | | | | | | | | KernelScale: 2.7758 | | 104 | 6 | Accept | 0.17255 | 191.76 | 0.14353 | 0.14571 | svm | BoxConstraint: 0.055759 | | | | | | | | | | KernelScale: 18.892 | | 105 | 6 | Accept | 0.15195 | 158.78 | 0.14353 | 0.14564 | svm | BoxConstraint: 0.012192 | | | | | | | | | | KernelScale: 3.9397 | | 106 | 6 | Accept | 0.16752 | 186.75 | 0.14353 | 0.14569 | svm | BoxConstraint: 0.074389 | | | | | | | | | | KernelScale: 18.359 | | 107 | 6 | Accept | 0.14911 | 153.52 | 0.14353 | 0.14596 | svm | BoxConstraint: 0.011498 | | | | | | | | | | KernelScale: 3.1239 | | 108 | 6 | Accept | 0.22432 | 3370.4 | 0.14353 | 0.14617 | svm | BoxConstraint: 0.26224 | | | | | | | | | | KernelScale: 0.48333 | | 109 | 6 | Accept | 0.14851 | 150.42 | 0.14353 | 0.14601 | svm | BoxConstraint: 0.011062 | | | | | | | | | | KernelScale: 2.9993 | | 110 | 6 | Accept | 0.14605 | 302.86 | 0.14353 | 0.14605 | svm | BoxConstraint: 1.2847 | | | | | | | | | | KernelScale: 3.6109 | | 111 | 6 | Accept | 0.14905 | 479.09 | 0.14353 | 0.14572 | svm | BoxConstraint: 0.37863 | | | | | | | | | | KernelScale: 2.2975 | | 112 | 6 | Accept | 0.17238 | 188.74 | 0.14353 | 0.14591 | svm | BoxConstraint: 0.011768 | | | | | | | | | | KernelScale: 9.0656 | | 113 | 6 | Accept | 0.14986 | 154.37 | 0.14353 | 0.1458 | svm | BoxConstraint: 0.0091472 | | | | | | | | | | KernelScale: 3.0854 | | 114 | 6 | Accept | 0.14945 | 150.17 | 0.14353 | 0.1456 | svm | BoxConstraint: 0.010472 | | | | | | | | | | KernelScale: 3.1579 | | 115 | 6 | Accept | 0.14585 | 336.82 | 0.14353 | 0.14577 | svm | BoxConstraint: 2.477 | | | | | | | | | | KernelScale: 4.0884 | | 116 | 6 | Accept | 0.14579 | 226.31 | 0.14353 | 0.14583 | svm | BoxConstraint: 0.19394 | | | | | | | | | | KernelScale: 2.6603 | | 117 | 6 | Accept | 0.15424 | 157.96 | 0.14353 | 0.14587 | svm | BoxConstraint: 0.33622 | | | | | | | | | | KernelScale: 18.198 | | 118 | 6 | Accept | 0.14706 | 311.23 | 0.14353 | 0.14599 | svm | BoxConstraint: 0.27404 | | | | | | | | | | KernelScale: 2.4769 | | 119 | 6 | Accept | 0.15064 | 576.64 | 0.14353 | 0.1456 | svm | BoxConstraint: 0.20057 | | | | | | | | | | KernelScale: 1.9187 | | 120 | 6 | Accept | 0.22663 | 3502.9 | 0.14353 | 0.14568 | svm | BoxConstraint: 26.696 | | | | | | | | | | KernelScale: 0.83468 |
__________________________________________________________ Optimization completed. Total iterations: 120 Total elapsed time: 11529.2732 seconds Total time for training and validation: 61054.1309 seconds Best observed learner is a tree model with: Learner: tree MinLeafSize: 79 Observed validation loss: 0.14353 Time for training and validation: 1.687 seconds Best estimated learner (returned model) is an svm model with: Learner: svm BoxConstraint: 0.29922 KernelScale: 4.0302 Estimated validation loss: 0.14568 Estimated time for training and validation: 168.4716 seconds Documentation for fitcauto display
Total elapsed time の値から、ベイズ最適化の実行に時間がかかったことがわかります (約 3.2 時間)。
fitcauto によって返される最終的なモデルが、最適な推定学習器となります。モデルを返す前に、関数は学習データ セット全体 (adultdata)、リストされている Learner (またはモデル) のタイプ、および表示されたハイパーパラメーター値を使用して、モデルの再学習を行います。
ASHA 最適化による自動モデル選択の使用
学習セットの観測値の数が原因でベイズ最適化による fitcauto の実行に長い時間がかかる場合は、代わりに ASHA 最適化による fitcauto を使用することを検討してください。adultdata に含まれる観測値が 10,000 を超える場合は、ASHA 最適化による fitcauto を使用して適切な分類器を自動的に見つけるよう試します。ASHA 最適化による fitcauto を使用すると、関数はさまざまなハイパーパラメーターの値をもつ複数のモデルを無作為に選択し、学習データの小さいサブセットで学習させます。交差検証の分類誤差 (Validation Loss) に基づいて有望なモデルが見つかると、そのモデルをプロモートし、より多くの学習データで学習させます。このプロセスを繰り返し、データの量を徐々に増やしながら有望なモデルに学習させます。既定の設定では、fitcauto は、最適化のプロット、および最適化の結果の反復表示を提供します。これらの結果を解釈する方法の詳細については、Verbose の表示を参照してください。
観測値の重みを設定し、ASHA 最適化を並列実行するよう指定します。ASHA 最適化は既定のベイズ最適化に比べて反復回数が多くなる場合が多いことに注意してください。時間の制約がある場合は、HyperparameterOptimizationOptions 構造体の MaxTime フィールドを指定して、fitcauto を実行する秒数を制限できます。
ashaOptions = struct("Optimizer","asha","UseParallel",true); [ashaMdl,ashaResults] = fitcauto(adultdata,"salary","Weights","fnlwgt", ... "HyperparameterOptimizationOptions",ashaOptions);
Copying objective function to workers...
Warning: Files that have already been attached are being ignored. To see which files are attached see the 'AttachedFiles' property of the parallel pool.
Done copying objective function to workers. Learner types to explore: ensemble, nb, svm, tree Total iterations (MaxObjectiveEvaluations): 425 Total time (MaxTime): Inf |====================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Training set | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | size | | | |====================================================================================================================================================| | 1 | 6 | Best | 0.24677 | 0.41223 | 0.24677 | 378 | tree | MinLeafSize: 293 | | 2 | 5 | Best | 0.21242 | 0.69352 | 0.21242 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 3 | 5 | Accept | 0.24677 | 0.49992 | 0.21242 | 378 | tree | MinLeafSize: 1192 | | 4 | 5 | Accept | 0.24677 | 0.89242 | 0.21242 | 378 | tree | MinLeafSize: 10141 | | 5 | 5 | Best | 0.18925 | 0.26458 | 0.18925 | 378 | tree | MinLeafSize: 16 | | 6 | 6 | Best | 0.18343 | 0.41943 | 0.18343 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 7 | 6 | Accept | 0.24677 | 0.98959 | 0.18343 | 378 | svm | BoxConstraint: 1.3947 | | | | | | | | | | KernelScale: 65.678 | | 8 | 6 | Accept | 0.18372 | 0.53536 | 0.18343 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 9 | 6 | Accept | 0.1954 | 9.5036 | 0.18343 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.0034879 | | | | | | | | | | Standardize: true | | 10 | 6 | Best | 0.1792 | 0.5341 | 0.1792 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 11 | 6 | Best | 0.17449 | 0.425 | 0.17449 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 12 | 6 | Accept | 0.24677 | 14.424 | 0.17449 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 295 | | | | | | | | | | MinLeafSize: 4119 | | 13 | 6 | Accept | 0.20024 | 0.42798 | 0.17449 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 14 | 6 | Accept | 0.20304 | 13.33 | 0.17449 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 2.8059 | | | | | | | | | | Standardize: true | | 15 | 6 | Accept | 0.19474 | 14.061 | 0.17449 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 25.229 | | | | | | | | | | Standardize: true | | 16 | 6 | Best | 0.16052 | 0.3778 | 0.16052 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 17 | 6 | Accept | 0.20451 | 0.20402 | 0.16052 | 378 | tree | MinLeafSize: 19 | | 18 | 6 | Accept | 0.24677 | 0.1917 | 0.16052 | 378 | tree | MinLeafSize: 1648 | | 19 | 6 | Accept | 0.182 | 0.48008 | 0.16052 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 20 | 6 | Accept | 0.21976 | 0.84268 | 0.16052 | 378 | svm | BoxConstraint: 1.374 | | | | | | | | | | KernelScale: 1.1138 | |====================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Training set | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | size | | | |====================================================================================================================================================| | 21 | 6 | Accept | 0.16445 | 0.40264 | 0.16052 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 22 | 6 | Accept | 0.17111 | 0.39243 | 0.16052 | 6033 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 23 | 6 | Accept | 0.17395 | 0.87746 | 0.16052 | 378 | svm | BoxConstraint: 1.3291 | | | | | | | | | | KernelScale: 11.5 | | 24 | 6 | Accept | 0.24677 | 0.88712 | 0.16052 | 378 | svm | BoxConstraint: 2.32 | | | | | | | | | | KernelScale: 249.93 | | 25 | 6 | Accept | 0.18863 | 8.5476 | 0.16052 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.050938 | | | | | | | | | | Standardize: true | | 26 | 6 | Accept | 0.39894 | 19.121 | 0.16052 | 378 | svm | BoxConstraint: 0.082795 | | | | | | | | | | KernelScale: 0.11799 | | 27 | 6 | Accept | 0.21388 | 0.84483 | 0.16052 | 378 | svm | BoxConstraint: 536.31 | | | | | | | | | | KernelScale: 2.96 | | 28 | 6 | Accept | 0.17673 | 0.40821 | 0.16052 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 29 | 6 | Accept | 0.24677 | 0.14927 | 0.16052 | 378 | tree | MinLeafSize: 228 | | 30 | 6 | Best | 0.15579 | 2.158 | 0.15579 | 1509 | svm | BoxConstraint: 1.3291 | | | | | | | | | | KernelScale: 11.5 | | 31 | 6 | Accept | 0.16985 | 0.38832 | 0.15579 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 32 | 6 | Accept | 0.24677 | 0.30699 | 0.15579 | 378 | svm | BoxConstraint: 0.045199 | | | | | | | | | | KernelScale: 0.0063948 | | 33 | 6 | Accept | 0.20589 | 0.20066 | 0.15579 | 378 | tree | MinLeafSize: 22 | | 34 | 6 | Accept | 0.24677 | 0.90412 | 0.15579 | 378 | svm | BoxConstraint: 0.059749 | | | | | | | | | | KernelScale: 179.59 | | 35 | 6 | Accept | 0.17167 | 11.632 | 0.15579 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.3336 | | | | | | | | | | Standardize: true | | 36 | 6 | Accept | 0.22055 | 12.98 | 0.15579 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 12.182 | | | | | | | | | | Standardize: true | | 37 | 6 | Accept | 0.16867 | 12.841 | 0.15579 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 108.9 | | | | | | | | | | Standardize: false | | 38 | 6 | Accept | 0.16266 | 0.86177 | 0.15579 | 378 | svm | BoxConstraint: 385.35 | | | | | | | | | | KernelScale: 65.814 | | 39 | 6 | Accept | 0.22491 | 0.8454 | 0.15579 | 378 | svm | BoxConstraint: 11.832 | | | | | | | | | | KernelScale: 1.7245 | | 40 | 6 | Accept | 0.1561 | 2.1629 | 0.15579 | 1509 | svm | BoxConstraint: 385.35 | | | | | | | | | | KernelScale: 65.814 | |====================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Training set | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | size | | | |====================================================================================================================================================| | 41 | 6 | Accept | 0.22074 | 0.28317 | 0.15579 | 378 | tree | MinLeafSize: 22 | | 42 | 6 | Accept | 0.24677 | 12.744 | 0.15579 | 378 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 297 | | | | | | | | | | MinLeafSize: 7579 | | 43 | 6 | Accept | 0.24677 | 14.118 | 0.15579 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 282 | | | | | | | | | | MinLeafSize: 1543 | | 44 | 6 | Accept | 0.16998 | 0.33582 | 0.15579 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 45 | 6 | Accept | 0.31094 | 37.128 | 0.15579 | 378 | svm | BoxConstraint: 70.951 | | | | | | | | | | KernelScale: 0.033775 | | 46 | 6 | Accept | 0.24677 | 0.34815 | 0.15579 | 378 | svm | BoxConstraint: 623.48 | | | | | | | | | | KernelScale: 0.0010192 | | 47 | 6 | Accept | 0.17045 | 10.999 | 0.15579 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 1.8198 | | | | | | | | | | Standardize: false | | 48 | 6 | Accept | 0.15807 | 13.413 | 0.15579 | 378 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 220 | | | | | | | | | | MinLeafSize: 30 | | 49 | 6 | Accept | 0.17396 | 29.339 | 0.15579 | 1509 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.3336 | | | | | | | | | | Standardize: true | | 50 | 6 | Accept | 0.676 | 23.204 | 0.15579 | 378 | svm | BoxConstraint: 39.423 | | | | | | | | | | KernelScale: 0.10596 | | 51 | 6 | Accept | 0.20495 | 0.22109 | 0.15579 | 378 | tree | MinLeafSize: 21 | | 52 | 6 | Accept | 0.18563 | 0.20139 | 0.15579 | 378 | tree | MinLeafSize: 42 | | 53 | 6 | Accept | 0.24677 | 0.92178 | 0.15579 | 378 | svm | BoxConstraint: 0.05188 | | | | | | | | | | KernelScale: 44.452 | | 54 | 6 | Accept | 0.18321 | 0.38725 | 0.15579 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 55 | 6 | Accept | 0.17723 | 0.4152 | 0.15579 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 56 | 6 | Accept | 0.48003 | 26.08 | 0.15579 | 378 | svm | BoxConstraint: 213.45 | | | | | | | | | | KernelScale: 0.092072 | | 57 | 6 | Best | 0.15393 | 17.242 | 0.15393 | 1509 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 220 | | | | | | | | | | MinLeafSize: 30 | | 58 | 6 | Accept | 0.17298 | 0.89838 | 0.15393 | 378 | svm | BoxConstraint: 962.26 | | | | | | | | | | KernelScale: 207.85 | | 59 | 6 | Best | 0.153 | 12.348 | 0.153 | 6033 | svm | BoxConstraint: 1.3291 | | | | | | | | | | KernelScale: 11.5 | | 60 | 6 | Accept | 0.18192 | 0.89064 | 0.153 | 378 | svm | BoxConstraint: 0.0055891 | | | | | | | | | | KernelScale: 0.90898 | |====================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Training set | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | size | | | |====================================================================================================================================================| | 61 | 6 | Accept | 0.18271 | 32.278 | 0.153 | 1509 | nb | DistributionNames: kernel | | | | | | | | | | Width: 108.9 | | | | | | | | | | Standardize: false | | 62 | 6 | Accept | 0.24677 | 0.14112 | 0.153 | 378 | tree | MinLeafSize: 3783 | | 63 | 6 | Accept | 0.26476 | 34.581 | 0.153 | 378 | svm | BoxConstraint: 74.583 | | | | | | | | | | KernelScale: 0.059475 | | 64 | 6 | Accept | 0.20762 | 8.8087 | 0.153 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.10297 | | | | | | | | | | Standardize: false | | 65 | 6 | Accept | 0.24677 | 1.0071 | 0.153 | 378 | svm | BoxConstraint: 0.09297 | | | | | | | | | | KernelScale: 536.28 | | 66 | 6 | Accept | 0.24194 | 1.071 | 0.153 | 378 | svm | BoxConstraint: 0.67333 | | | | | | | | | | KernelScale: 35.903 | | 67 | 6 | Accept | 0.20385 | 18.897 | 0.153 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 252 | | | | | | | | | | MinLeafSize: 77 | | 68 | 6 | Accept | 0.15503 | 2.2706 | 0.153 | 1509 | svm | BoxConstraint: 962.26 | | | | | | | | | | KernelScale: 207.85 | | 69 | 6 | Accept | 0.24677 | 13.432 | 0.153 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 245 | | | | | | | | | | MinLeafSize: 3552 | | 70 | 6 | Accept | 0.19077 | 8.1226 | 0.153 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.027553 | | | | | | | | | | Standardize: false | | 71 | 6 | Accept | 0.16757 | 0.36842 | 0.153 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 72 | 6 | Accept | 0.18779 | 0.35808 | 0.153 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 73 | 6 | Accept | 0.17793 | 12.899 | 0.153 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 21.131 | | | | | | | | | | Standardize: false | | 74 | 6 | Accept | 0.16496 | 27.002 | 0.153 | 1509 | nb | DistributionNames: kernel | | | | | | | | | | Width: 1.8198 | | | | | | | | | | Standardize: false | | 75 | 6 | Accept | 0.24677 | 0.21047 | 0.153 | 378 | tree | MinLeafSize: 12519 | | 76 | 6 | Accept | 0.16784 | 0.90871 | 0.153 | 378 | svm | BoxConstraint: 763.29 | | | | | | | | | | KernelScale: 144.74 | | 77 | 6 | Accept | 0.24677 | 8.515 | 0.153 | 378 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 210 | | | | | | | | | | MinLeafSize: 604 | | 78 | 6 | Accept | 0.15451 | 2.0399 | 0.153 | 1509 | svm | BoxConstraint: 763.29 | | | | | | | | | | KernelScale: 144.74 | | 79 | 6 | Accept | 0.61916 | 18.894 | 0.153 | 378 | svm | BoxConstraint: 0.043267 | | | | | | | | | | KernelScale: 0.14745 | | 80 | 6 | Accept | 0.17582 | 0.3839 | 0.153 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | |====================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Training set | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | size | | | |====================================================================================================================================================| | 81 | 6 | Accept | 0.24677 | 0.93542 | 0.153 | 378 | svm | BoxConstraint: 0.024337 | | | | | | | | | | KernelScale: 76.684 | | 82 | 6 | Accept | 0.17903 | 0.39209 | 0.153 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 83 | 6 | Accept | 0.26744 | 38.332 | 0.153 | 378 | svm | BoxConstraint: 171.89 | | | | | | | | | | KernelScale: 0.0079675 | | 84 | 6 | Accept | 0.20799 | 0.27606 | 0.153 | 378 | tree | MinLeafSize: 20 | | 85 | 6 | Accept | 0.24677 | 12.493 | 0.153 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 225 | | | | | | | | | | MinLeafSize: 8606 | | 86 | 6 | Accept | 0.24677 | 1.0149 | 0.153 | 378 | svm | BoxConstraint: 0.038237 | | | | | | | | | | KernelScale: 63.926 | | 87 | 6 | Accept | 0.24677 | 11.714 | 0.153 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 203 | | | | | | | | | | MinLeafSize: 332 | | 88 | 6 | Best | 0.15053 | 28.585 | 0.15053 | 6033 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 220 | | | | | | | | | | MinLeafSize: 30 | | 89 | 6 | Accept | 0.24677 | 0.19642 | 0.15053 | 378 | tree | MinLeafSize: 683 | | 90 | 6 | Accept | 0.24677 | 14.949 | 0.15053 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 241 | | | | | | | | | | MinLeafSize: 211 | | 91 | 6 | Accept | 0.24677 | 0.97254 | 0.15053 | 378 | svm | BoxConstraint: 0.0035269 | | | | | | | | | | KernelScale: 117.31 | | 92 | 6 | Best | 0.15028 | 13.847 | 0.15028 | 6033 | svm | BoxConstraint: 763.29 | | | | | | | | | | KernelScale: 144.74 | | 93 | 6 | Accept | 0.18829 | 2.5565 | 0.15028 | 1509 | svm | BoxConstraint: 0.0055891 | | | | | | | | | | KernelScale: 0.90898 | | 94 | 6 | Accept | 0.24677 | 12.589 | 0.15028 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 205 | | | | | | | | | | MinLeafSize: 4488 | | 95 | 6 | Accept | 0.19257 | 10.01 | 0.15028 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.92919 | | | | | | | | | | Standardize: false | | 96 | 6 | Accept | 0.1724 | 0.3932 | 0.15028 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 97 | 6 | Accept | 0.24677 | 13.185 | 0.15028 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 251 | | | | | | | | | | MinLeafSize: 5425 | | 98 | 6 | Accept | 0.18466 | 0.41239 | 0.15028 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 99 | 6 | Accept | 0.24677 | 0.15762 | 0.15028 | 378 | tree | MinLeafSize: 632 | | 100 | 6 | Accept | 0.18538 | 0.40593 | 0.15028 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | |====================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Training set | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | size | | | |====================================================================================================================================================| | 101 | 6 | Accept | 0.24735 | 0.1677 | 0.15028 | 378 | tree | MinLeafSize: 159 | | 102 | 6 | Accept | 0.46239 | 19.011 | 0.15028 | 378 | svm | BoxConstraint: 622.29 | | | | | | | | | | KernelScale: 0.13209 | | 103 | 6 | Accept | 0.17613 | 0.43685 | 0.15028 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 104 | 6 | Accept | 0.45428 | 21.612 | 0.15028 | 378 | svm | BoxConstraint: 243.32 | | | | | | | | | | KernelScale: 0.11649 | | 105 | 6 | Accept | 0.17695 | 33.598 | 0.15028 | 1509 | nb | DistributionNames: kernel | | | | | | | | | | Width: 21.131 | | | | | | | | | | Standardize: false | | 106 | 6 | Accept | 0.20001 | 0.89752 | 0.15028 | 378 | svm | BoxConstraint: 0.13074 | | | | | | | | | | KernelScale: 4.2955 | | 107 | 6 | Accept | 0.26791 | 34.811 | 0.15028 | 378 | svm | BoxConstraint: 0.15408 | | | | | | | | | | KernelScale: 0.0081934 | | 108 | 6 | Accept | 0.24677 | 12.075 | 0.15028 | 378 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 277 | | | | | | | | | | MinLeafSize: 4646 | | 109 | 6 | Accept | 0.17992 | 0.87394 | 0.15028 | 378 | svm | BoxConstraint: 97.764 | | | | | | | | | | KernelScale: 20.653 | | 110 | 6 | Accept | 0.17688 | 0.38115 | 0.15028 | 1509 | tree | MinLeafSize: 42 | | 111 | 6 | Accept | 0.1517 | 13.832 | 0.15028 | 6033 | svm | BoxConstraint: 962.26 | | | | | | | | | | KernelScale: 207.85 | | 112 | 6 | Accept | 0.18863 | 0.4304 | 0.15028 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 113 | 6 | Accept | 0.18648 | 0.453 | 0.15028 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 114 | 6 | Accept | 0.15876 | 19.737 | 0.15028 | 378 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 271 | | | | | | | | | | MinLeafSize: 14 | | 115 | 6 | Accept | 0.24184 | 0.90896 | 0.15028 | 378 | svm | BoxConstraint: 0.26232 | | | | | | | | | | KernelScale: 13.726 | | 116 | 6 | Accept | 0.16007 | 16.574 | 0.15028 | 378 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 226 | | | | | | | | | | MinLeafSize: 1 | | 117 | 6 | Accept | 0.18577 | 0.43107 | 0.15028 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 118 | 6 | Accept | 0.24677 | 0.2089 | 0.15028 | 378 | tree | MinLeafSize: 112 | | 119 | 6 | Accept | 0.16662 | 23.782 | 0.15028 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 291 | | | | | | | | | | MinLeafSize: 20 | | 120 | 6 | Accept | 0.18481 | 0.3758 | 0.15028 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | |====================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Training set | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | size | | | |====================================================================================================================================================| | 121 | 6 | Accept | 0.21737 | 13.332 | 0.15028 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 50.518 | | | | | | | | | | Standardize: true | | 122 | 6 | Accept | 0.24677 | 0.92059 | 0.15028 | 378 | svm | BoxConstraint: 0.0042093 | | | | | | | | | | KernelScale: 11.241 | | 123 | 6 | Accept | 0.15316 | 17.743 | 0.15028 | 1509 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 226 | | | | | | | | | | MinLeafSize: 1 | | 124 | 6 | Accept | 0.24677 | 0.16148 | 0.15028 | 378 | tree | MinLeafSize: 302 | | 125 | 6 | Accept | 0.15316 | 21.59 | 0.15028 | 1509 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 271 | | | | | | | | | | MinLeafSize: 14 | | 126 | 6 | Accept | 0.23141 | 37.152 | 0.15028 | 378 | svm | BoxConstraint: 21.777 | | | | | | | | | | KernelScale: 0.029276 | | 127 | 6 | Accept | 0.23304 | 16.815 | 0.15028 | 378 | svm | BoxConstraint: 173.49 | | | | | | | | | | KernelScale: 0.21811 | | 128 | 6 | Accept | 0.24677 | 0.142 | 0.15028 | 378 | tree | MinLeafSize: 2502 | | 129 | 6 | Accept | 0.15706 | 2.1789 | 0.15028 | 1509 | svm | BoxConstraint: 97.764 | | | | | | | | | | KernelScale: 20.653 | | 130 | 6 | Accept | 0.2435 | 12.745 | 0.15028 | 378 | svm | BoxConstraint: 0.053634 | | | | | | | | | | KernelScale: 0.31732 | | 131 | 6 | Accept | 0.21974 | 13.616 | 0.15028 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 3.3625 | | | | | | | | | | Standardize: true | | 132 | 6 | Accept | 0.19986 | 0.42586 | 0.15028 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 133 | 6 | Accept | 0.24677 | 0.33075 | 0.15028 | 378 | svm | BoxConstraint: 466.95 | | | | | | | | | | KernelScale: 0.0044945 | | 134 | 6 | Accept | 0.16879 | 0.40206 | 0.15028 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 135 | 6 | Accept | 0.24677 | 0.1607 | 0.15028 | 378 | tree | MinLeafSize: 724 | | 136 | 6 | Accept | 0.15942 | 27.353 | 0.15028 | 1509 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 291 | | | | | | | | | | MinLeafSize: 20 | | 137 | 6 | Accept | 0.1828 | 0.43307 | 0.15028 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 138 | 6 | Accept | 0.24677 | 11.762 | 0.15028 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 212 | | | | | | | | | | MinLeafSize: 2353 | | 139 | 6 | Accept | 0.24677 | 14.322 | 0.15028 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 255 | | | | | | | | | | MinLeafSize: 2166 | | 140 | 6 | Accept | 0.19199 | 0.41275 | 0.15028 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | |====================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Training set | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | size | | | |====================================================================================================================================================| | 141 | 6 | Accept | 0.18811 | 0.37075 | 0.15028 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 142 | 6 | Accept | 0.24677 | 9.4582 | 0.15028 | 378 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 203 | | | | | | | | | | MinLeafSize: 3125 | | 143 | 6 | Accept | 0.24189 | 0.97907 | 0.15028 | 378 | svm | BoxConstraint: 2.2181 | | | | | | | | | | KernelScale: 60.618 | | 144 | 6 | Accept | 0.22596 | 0.81415 | 0.15028 | 378 | svm | BoxConstraint: 8.9199 | | | | | | | | | | KernelScale: 0.93704 | | 145 | 6 | Accept | 0.17702 | 0.43906 | 0.15028 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 146 | 6 | Accept | 0.24749 | 33.011 | 0.15028 | 378 | svm | BoxConstraint: 0.0023424 | | | | | | | | | | KernelScale: 0.02797 | | 147 | 6 | Accept | 0.24677 | 0.3189 | 0.15028 | 378 | svm | BoxConstraint: 0.59521 | | | | | | | | | | KernelScale: 0.0066679 | | 148 | 6 | Accept | 0.24591 | 13.225 | 0.15028 | 378 | svm | BoxConstraint: 521.24 | | | | | | | | | | KernelScale: 0.28175 | | 149 | 6 | Accept | 0.22288 | 0.91802 | 0.15028 | 378 | svm | BoxConstraint: 0.021474 | | | | | | | | | | KernelScale: 4.5183 | | 150 | 6 | Accept | 0.18334 | 0.40628 | 0.15028 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 151 | 6 | Accept | 0.18334 | 13.603 | 0.15028 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 693.96 | | | | | | | | | | Standardize: false | | 152 | 6 | Best | 0.14984 | 34.6 | 0.14984 | 6033 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 271 | | | | | | | | | | MinLeafSize: 14 | | 153 | 6 | Accept | 0.24677 | 0.93891 | 0.14984 | 378 | svm | BoxConstraint: 0.011464 | | | | | | | | | | KernelScale: 887.95 | | 154 | 6 | Accept | 0.24677 | 15.021 | 0.14984 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 271 | | | | | | | | | | MinLeafSize: 3858 | | 155 | 6 | Accept | 0.24677 | 0.2819 | 0.14984 | 378 | svm | BoxConstraint: 0.017 | | | | | | | | | | KernelScale: 0.0046035 | | 156 | 6 | Accept | 0.16761 | 10.948 | 0.14984 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 1.613 | | | | | | | | | | Standardize: false | | 157 | 6 | Accept | 0.24677 | 0.30553 | 0.14984 | 378 | svm | BoxConstraint: 4.1988 | | | | | | | | | | KernelScale: 0.0011787 | | 158 | 6 | Accept | 0.17316 | 22.007 | 0.14984 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 276 | | | | | | | | | | MinLeafSize: 16 | | 159 | 6 | Accept | 0.20481 | 0.86066 | 0.14984 | 378 | svm | BoxConstraint: 0.01545 | | | | | | | | | | KernelScale: 1.0258 | | 160 | 6 | Accept | 0.15008 | 29.64 | 0.14984 | 6033 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 226 | | | | | | | | | | MinLeafSize: 1 | |====================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Training set | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | size | | | |====================================================================================================================================================| | 161 | 6 | Accept | 0.18437 | 0.9387 | 0.14984 | 378 | svm | BoxConstraint: 0.0087948 | | | | | | | | | | KernelScale: 1.1494 | | 162 | 6 | Accept | 0.20904 | 13.569 | 0.14984 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 91.024 | | | | | | | | | | Standardize: true | | 163 | 6 | Accept | 0.23451 | 6.5741 | 0.14984 | 378 | svm | BoxConstraint: 0.11495 | | | | | | | | | | KernelScale: 0.34761 | | 164 | 6 | Best | 0.1493 | 147.7 | 0.1493 | 24130 | svm | BoxConstraint: 763.29 | | | | | | | | | | KernelScale: 144.74 | | 165 | 6 | Accept | 0.27656 | 40.771 | 0.1493 | 378 | svm | BoxConstraint: 1.0475 | | | | | | | | | | KernelScale: 0.10172 | | 166 | 6 | Accept | 0.16234 | 26.581 | 0.1493 | 1509 | nb | DistributionNames: kernel | | | | | | | | | | Width: 1.613 | | | | | | | | | | Standardize: false | | 167 | 6 | Accept | 0.19497 | 0.19102 | 0.1493 | 378 | tree | MinLeafSize: 57 | | 168 | 6 | Accept | 0.19417 | 34.123 | 0.1493 | 1509 | nb | DistributionNames: kernel | | | | | | | | | | Width: 693.96 | | | | | | | | | | Standardize: false | | 169 | 6 | Accept | 0.25929 | 19.29 | 0.1493 | 378 | svm | BoxConstraint: 19.843 | | | | | | | | | | KernelScale: 0.14367 | | 170 | 6 | Accept | 0.162 | 0.84268 | 0.1493 | 378 | svm | BoxConstraint: 13.978 | | | | | | | | | | KernelScale: 30.295 | | 171 | 6 | Accept | 0.16933 | 0.83709 | 0.1493 | 378 | svm | BoxConstraint: 3.9363 | | | | | | | | | | KernelScale: 17.816 | | 172 | 6 | Accept | 0.15652 | 2.1022 | 0.1493 | 1509 | svm | BoxConstraint: 13.978 | | | | | | | | | | KernelScale: 30.295 | | 173 | 6 | Accept | 0.22197 | 12.696 | 0.1493 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 102.13 | | | | | | | | | | Standardize: true | | 174 | 6 | Accept | 0.24677 | 0.18644 | 0.1493 | 378 | tree | MinLeafSize: 864 | | 175 | 6 | Accept | 0.24677 | 0.15003 | 0.1493 | 378 | tree | MinLeafSize: 1322 | | 176 | 6 | Accept | 0.18647 | 8.0636 | 0.1493 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.013221 | | | | | | | | | | Standardize: true | | 177 | 6 | Accept | 0.15981 | 2.1353 | 0.1493 | 1509 | svm | BoxConstraint: 3.9363 | | | | | | | | | | KernelScale: 17.816 | | 178 | 6 | Accept | 0.18089 | 0.43329 | 0.1493 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 179 | 6 | Accept | 0.17375 | 0.35426 | 0.1493 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 180 | 6 | Accept | 0.16252 | 0.38204 | 0.1493 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | |====================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Training set | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | size | | | |====================================================================================================================================================| | 181 | 6 | Accept | 0.38861 | 33.253 | 0.1493 | 378 | svm | BoxConstraint: 6.5307 | | | | | | | | | | KernelScale: 0.072466 | | 182 | 6 | Accept | 0.17638 | 0.43308 | 0.1493 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 183 | 6 | Accept | 0.15072 | 13.541 | 0.1493 | 6033 | svm | BoxConstraint: 385.35 | | | | | | | | | | KernelScale: 65.814 | | 184 | 6 | Accept | 0.15577 | 25.223 | 0.1493 | 1509 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 276 | | | | | | | | | | MinLeafSize: 16 | | 185 | 6 | Accept | 0.24677 | 14.695 | 0.1493 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 270 | | | | | | | | | | MinLeafSize: 1304 | | 186 | 6 | Accept | 0.19237 | 0.42507 | 0.1493 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 187 | 6 | Accept | 0.19325 | 0.38775 | 0.1493 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 188 | 6 | Accept | 0.19327 | 9.6642 | 0.1493 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.58748 | | | | | | | | | | Standardize: false | | 189 | 6 | Accept | 0.18255 | 13.18 | 0.1493 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 54.029 | | | | | | | | | | Standardize: false | | 190 | 6 | Accept | 0.16223 | 0.33709 | 0.1493 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 191 | 6 | Accept | 0.19958 | 0.17893 | 0.1493 | 378 | tree | MinLeafSize: 24 | | 192 | 6 | Accept | 0.24677 | 0.9243 | 0.1493 | 378 | svm | BoxConstraint: 0.028158 | | | | | | | | | | KernelScale: 21.174 | | 193 | 6 | Accept | 0.18379 | 9.5712 | 0.1493 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.69661 | | | | | | | | | | Standardize: false | | 194 | 6 | Accept | 0.17841 | 0.41127 | 0.1493 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 195 | 6 | Accept | 0.49303 | 36.252 | 0.1493 | 378 | svm | BoxConstraint: 0.035455 | | | | | | | | | | KernelScale: 0.016789 | | 196 | 6 | Accept | 0.24677 | 0.12499 | 0.1493 | 378 | tree | MinLeafSize: 233 | | 197 | 6 | Accept | 0.24677 | 11.467 | 0.1493 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 209 | | | | | | | | | | MinLeafSize: 398 | | 198 | 6 | Accept | 0.17691 | 0.86872 | 0.1493 | 378 | svm | BoxConstraint: 596.59 | | | | | | | | | | KernelScale: 21.708 | | 199 | 6 | Accept | 0.15712 | 2.9033 | 0.1493 | 1509 | svm | BoxConstraint: 596.59 | | | | | | | | | | KernelScale: 21.708 | | 200 | 6 | Accept | 0.17137 | 0.3622 | 0.1493 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | |====================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Training set | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | size | | | |====================================================================================================================================================| | 201 | 6 | Accept | 0.24677 | 16.885 | 0.1493 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 299 | | | | | | | | | | MinLeafSize: 216 | | 202 | 6 | Accept | 0.24677 | 11.468 | 0.1493 | 378 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 213 | | | | | | | | | | MinLeafSize: 1448 | | 203 | 6 | Accept | 0.16841 | 1.0814 | 0.1493 | 378 | svm | BoxConstraint: 484.15 | | | | | | | | | | KernelScale: 69.555 | | 204 | 5 | Accept | 0.17402 | 13.714 | 0.1493 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 3.8082 | | | | | | | | | | Standardize: false | | 205 | 5 | Accept | 0.15642 | 2.2741 | 0.1493 | 1509 | svm | BoxConstraint: 484.15 | | | | | | | | | | KernelScale: 69.555 | | 206 | 6 | Accept | 0.24677 | 0.30604 | 0.1493 | 378 | svm | BoxConstraint: 0.022208 | | | | | | | | | | KernelScale: 0.0071215 | | 207 | 6 | Accept | 0.1817 | 0.42424 | 0.1493 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 208 | 6 | Accept | 0.18218 | 0.89991 | 0.1493 | 378 | svm | BoxConstraint: 162.42 | | | | | | | | | | KernelScale: 10.207 | | 209 | 6 | Accept | 0.24677 | 0.92031 | 0.1493 | 378 | svm | BoxConstraint: 28.319 | | | | | | | | | | KernelScale: 317.4 | | 210 | 6 | Accept | 0.17479 | 0.38641 | 0.1493 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 211 | 6 | Accept | 0.225 | 0.86838 | 0.1493 | 378 | svm | BoxConstraint: 810.07 | | | | | | | | | | KernelScale: 3.0301 | | 212 | 6 | Accept | 0.17704 | 0.40973 | 0.1493 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 213 | 6 | Accept | 0.19395 | 11.826 | 0.1493 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.27286 | | | | | | | | | | Standardize: true | | 214 | 6 | Accept | 0.17788 | 21.7 | 0.1493 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 283 | | | | | | | | | | MinLeafSize: 47 | | 215 | 6 | Accept | 0.14987 | 14.362 | 0.1493 | 6033 | svm | BoxConstraint: 484.15 | | | | | | | | | | KernelScale: 69.555 | | 216 | 6 | Accept | 0.21628 | 13.595 | 0.1493 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 9.774 | | | | | | | | | | Standardize: true | | 217 | 5 | Accept | 0.17398 | 0.77326 | 0.1493 | 378 | svm | BoxConstraint: 112.02 | | | | | | | | | | KernelScale: 13.915 | | 218 | 5 | Accept | 0.24677 | 0.1336 | 0.1493 | 378 | tree | MinLeafSize: 653 | | 219 | 6 | Accept | 0.15218 | 40.273 | 0.1493 | 6033 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 276 | | | | | | | | | | MinLeafSize: 16 | | 220 | 6 | Accept | 0.16554 | 0.37094 | 0.1493 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | |====================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Training set | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | size | | | |====================================================================================================================================================| | 221 | 6 | Accept | 0.15334 | 2.3799 | 0.1493 | 1509 | svm | BoxConstraint: 112.02 | | | | | | | | | | KernelScale: 13.915 | | 222 | 6 | Accept | 0.19759 | 8.0999 | 0.1493 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.0023918 | | | | | | | | | | Standardize: true | | 223 | 6 | Accept | 0.23132 | 0.77865 | 0.1493 | 378 | svm | BoxConstraint: 10.854 | | | | | | | | | | KernelScale: 0.90505 | | 224 | 6 | Accept | 0.18694 | 8.0651 | 0.1493 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.16744 | | | | | | | | | | Standardize: false | | 225 | 6 | Accept | 0.17874 | 0.37035 | 0.1493 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 226 | 6 | Accept | 0.24677 | 0.14836 | 0.1493 | 378 | tree | MinLeafSize: 12624 | | 227 | 6 | Accept | 0.24677 | 11.097 | 0.1493 | 378 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 257 | | | | | | | | | | MinLeafSize: 228 | | 228 | 6 | Accept | 0.18473 | 0.45468 | 0.1493 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 229 | 6 | Accept | 0.16683 | 16.222 | 0.1493 | 378 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 274 | | | | | | | | | | MinLeafSize: 52 | | 230 | 6 | Accept | 0.24677 | 0.89615 | 0.1493 | 378 | svm | BoxConstraint: 0.0023661 | | | | | | | | | | KernelScale: 236.67 | | 231 | 6 | Accept | 0.21503 | 0.27122 | 0.1493 | 378 | tree | MinLeafSize: 3 | | 232 | 6 | Accept | 0.16733 | 18.627 | 0.1493 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 253 | | | | | | | | | | MinLeafSize: 15 | | 233 | 6 | Accept | 0.16675 | 29.958 | 0.1493 | 1509 | nb | DistributionNames: kernel | | | | | | | | | | Width: 3.8082 | | | | | | | | | | Standardize: false | | 234 | 6 | Accept | 0.16016 | 17.053 | 0.1493 | 378 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 261 | | | | | | | | | | MinLeafSize: 7 | | 235 | 6 | Accept | 0.16943 | 9.9624 | 0.1493 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 1.4142 | | | | | | | | | | Standardize: false | | 236 | 6 | Accept | 0.24677 | 0.18215 | 0.1493 | 378 | tree | MinLeafSize: 4920 | | 237 | 6 | Accept | 0.24677 | 11.257 | 0.1493 | 378 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 268 | | | | | | | | | | MinLeafSize: 2856 | | 238 | 6 | Accept | 0.21822 | 0.1878 | 0.1493 | 378 | tree | MinLeafSize: 62 | | 239 | 6 | Accept | 0.15342 | 21.65 | 0.1493 | 1509 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 274 | | | | | | | | | | MinLeafSize: 52 | | 240 | 6 | Accept | 0.19453 | 8.2977 | 0.1493 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.0053849 | | | | | | | | | | Standardize: true | |====================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Training set | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | size | | | |====================================================================================================================================================| | 241 | 6 | Best | 0.14836 | 96.23 | 0.14836 | 24130 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 271 | | | | | | | | | | MinLeafSize: 14 | | 242 | 6 | Accept | 0.24677 | 0.33161 | 0.14836 | 378 | svm | BoxConstraint: 346.19 | | | | | | | | | | KernelScale: 0.0031468 | | 243 | 6 | Accept | 0.15424 | 20.047 | 0.14836 | 1509 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 261 | | | | | | | | | | MinLeafSize: 7 | | 244 | 6 | Accept | 0.24677 | 0.18571 | 0.14836 | 378 | tree | MinLeafSize: 2558 | | 245 | 6 | Accept | 0.24677 | 0.35537 | 0.14836 | 378 | svm | BoxConstraint: 0.0016007 | | | | | | | | | | KernelScale: 0.0011917 | | 246 | 6 | Accept | 0.17468 | 11.674 | 0.14836 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.34418 | | | | | | | | | | Standardize: true | | 247 | 6 | Accept | 0.25459 | 33.254 | 0.14836 | 378 | svm | BoxConstraint: 492.71 | | | | | | | | | | KernelScale: 0.062372 | | 248 | 6 | Accept | 0.24677 | 0.1379 | 0.14836 | 378 | tree | MinLeafSize: 374 | | 249 | 6 | Accept | 0.15195 | 23.675 | 0.14836 | 1509 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 253 | | | | | | | | | | MinLeafSize: 15 | | 250 | 6 | Accept | 0.18005 | 10.469 | 0.14836 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.18024 | | | | | | | | | | Standardize: true | | 251 | 6 | Accept | 0.18113 | 10.289 | 0.14836 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.15646 | | | | | | | | | | Standardize: true | | 252 | 6 | Accept | 0.14951 | 23.305 | 0.14836 | 6033 | svm | BoxConstraint: 112.02 | | | | | | | | | | KernelScale: 13.915 | | 253 | 6 | Accept | 0.24661 | 0.17732 | 0.14836 | 378 | tree | MinLeafSize: 115 | | 254 | 6 | Accept | 0.24677 | 14.536 | 0.14836 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 275 | | | | | | | | | | MinLeafSize: 7160 | | 255 | 6 | Accept | 0.24677 | 0.14128 | 0.14836 | 378 | tree | MinLeafSize: 342 | | 256 | 6 | Accept | 0.17276 | 0.37834 | 0.14836 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 257 | 6 | Accept | 0.23016 | 12.927 | 0.14836 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 35.855 | | | | | | | | | | Standardize: true | | 258 | 6 | Accept | 0.24677 | 0.15532 | 0.14836 | 378 | tree | MinLeafSize: 7345 | | 259 | 6 | Accept | 0.24677 | 0.95577 | 0.14836 | 378 | svm | BoxConstraint: 0.020628 | | | | | | | | | | KernelScale: 164.3 | | 260 | 6 | Accept | 0.2025 | 0.18858 | 0.14836 | 378 | tree | MinLeafSize: 36 | |====================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Training set | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | size | | | |====================================================================================================================================================| | 261 | 6 | Accept | 0.16554 | 24.966 | 0.14836 | 1509 | nb | DistributionNames: kernel | | | | | | | | | | Width: 1.4142 | | | | | | | | | | Standardize: false | | 262 | 6 | Accept | 0.18069 | 10.339 | 0.14836 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.11577 | | | | | | | | | | Standardize: true | | 263 | 6 | Accept | 0.162 | 15.993 | 0.14836 | 378 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 245 | | | | | | | | | | MinLeafSize: 2 | | 264 | 6 | Accept | 0.24677 | 11.091 | 0.14836 | 378 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 266 | | | | | | | | | | MinLeafSize: 295 | | 265 | 6 | Accept | 0.20847 | 0.27559 | 0.14836 | 378 | tree | MinLeafSize: 2 | | 266 | 6 | Accept | 0.1769 | 30.334 | 0.14836 | 1509 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.34418 | | | | | | | | | | Standardize: true | | 267 | 6 | Accept | 0.24677 | 0.92783 | 0.14836 | 378 | svm | BoxConstraint: 1.8318 | | | | | | | | | | KernelScale: 640.96 | | 268 | 6 | Accept | 0.18559 | 0.38512 | 0.14836 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 269 | 6 | Accept | 0.17281 | 17.472 | 0.14836 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 251 | | | | | | | | | | MinLeafSize: 36 | | 270 | 6 | Accept | 0.24677 | 0.13322 | 0.14836 | 378 | tree | MinLeafSize: 1065 | | 271 | 6 | Accept | 0.23467 | 33.997 | 0.14836 | 378 | svm | BoxConstraint: 40.942 | | | | | | | | | | KernelScale: 0.060508 | | 272 | 6 | Accept | 0.16424 | 0.40692 | 0.14836 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 273 | 6 | Accept | 0.16911 | 25.586 | 0.14836 | 1509 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 283 | | | | | | | | | | MinLeafSize: 47 | | 274 | 6 | Accept | 0.17906 | 0.39597 | 0.14836 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 275 | 6 | Accept | 0.18946 | 8.6561 | 0.14836 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.0097303 | | | | | | | | | | Standardize: true | | 276 | 6 | Accept | 0.17244 | 0.39582 | 0.14836 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 277 | 6 | Accept | 0.19106 | 0.17636 | 0.14836 | 378 | tree | MinLeafSize: 58 | | 278 | 6 | Accept | 0.15421 | 18.602 | 0.14836 | 1509 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 245 | | | | | | | | | | MinLeafSize: 2 | | 279 | 6 | Accept | 0.1515 | 34.698 | 0.14836 | 6033 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 253 | | | | | | | | | | MinLeafSize: 15 | | 280 | 6 | Accept | 0.18897 | 13.381 | 0.14836 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 697.61 | | | | | | | | | | Standardize: true | |====================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Training set | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | size | | | |====================================================================================================================================================| | 281 | 6 | Accept | 0.16858 | 0.37927 | 0.14836 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 282 | 6 | Accept | 0.22522 | 13.462 | 0.14836 | 378 | svm | BoxConstraint: 0.07043 | | | | | | | | | | KernelScale: 0.25792 | | 283 | 6 | Accept | 0.2417 | 1.0039 | 0.14836 | 378 | svm | BoxConstraint: 0.0013375 | | | | | | | | | | KernelScale: 3.417 | | 284 | 6 | Accept | 0.18405 | 0.3932 | 0.14836 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 285 | 6 | Accept | 0.24677 | 0.33159 | 0.14836 | 378 | svm | BoxConstraint: 0.038245 | | | | | | | | | | KernelScale: 0.0021108 | | 286 | 6 | Accept | 0.19548 | 0.43701 | 0.14836 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 287 | 6 | Accept | 0.24677 | 1.0456 | 0.14836 | 378 | svm | BoxConstraint: 0.0050787 | | | | | | | | | | KernelScale: 664.76 | | 288 | 6 | Accept | 0.15713 | 0.41854 | 0.14836 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 289 | 6 | Accept | 0.24677 | 12.752 | 0.14836 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 232 | | | | | | | | | | MinLeafSize: 3957 | | 290 | 6 | Accept | 0.17541 | 0.38528 | 0.14836 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 291 | 6 | Accept | 0.16182 | 23.699 | 0.14836 | 1509 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 251 | | | | | | | | | | MinLeafSize: 36 | | 292 | 6 | Accept | 0.20209 | 8.1049 | 0.14836 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.036333 | | | | | | | | | | Standardize: false | | 293 | 6 | Accept | 0.24677 | 9.4156 | 0.14836 | 378 | ensemble | Method: LogitBoost | | | | | | | | | | NumLearningCycles: 208 | | | | | | | | | | MinLeafSize: 2039 | | 294 | 6 | Accept | 0.1795 | 0.41664 | 0.14836 | 1509 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 295 | 5 | Accept | 0.20244 | 8.2445 | 0.14836 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 0.003178 | | | | | | | | | | Standardize: false | | 296 | 5 | Accept | 0.22113 | 0.91625 | 0.14836 | 378 | svm | BoxConstraint: 83.258 | | | | | | | | | | KernelScale: 180.57 | | 297 | 6 | Accept | 0.24677 | 0.15149 | 0.14836 | 378 | tree | MinLeafSize: 330 | | 298 | 6 | Accept | 0.20546 | 0.23764 | 0.14836 | 378 | tree | MinLeafSize: 3 | | 299 | 6 | Accept | 0.17061 | 0.40961 | 0.14836 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | | 300 | 6 | Accept | 0.19405 | 0.41665 | 0.14836 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | | | Standardize: - | |====================================================================================================================================================| | Iter | Active | Eval | Validation | Time for training | Observed min | Training set | Learner | Hyperparameter: Value | | | workers | result | loss | & validation (sec)| validation loss | size | | | |====================================================================================================================================================| | 301 | 6 | Accept | 0.16126 | 17.363 | 0.14836 | 378 | ensemble | Method: Bag | | | | | | | | | | NumLearningCycles: 216 | | | | | | | | | | MinLeafSize: 2 | | 302 | 6 | Accept | 0.24677 | 0.1915 | 0.14836 | 378 | tree | MinLeafSize: 572 | | 303 | 6 | Accept | 0.20638 | 12.701 | 0.14836 | 378 | nb | DistributionNames: kernel | | | | | | | | | | Width: 4.8773 | | | | | | | | | | Standardize: false | | 304 | 6 | Accept | 0.15775 | 0.43674 | 0.14836 | 378 | nb | DistributionNames: normal | | | | | | | | | | Width: NaN | | | | | | | | ...
__________________________________________________________ Optimization completed. Total iterations: 425 Total elapsed time: 693.4747 seconds Total time for training and validation: 3784.6551 seconds Best observed learner is an ensemble model with: Learner: ensemble Method: LogitBoost NumLearningCycles: 271 MinLeafSize: 14 Observed validation loss: 0.14836 Time for training and validation: 96.2296 seconds Documentation for fitcauto display
Total elapsed time の値から、ASHA 最適化の方がベイズ最適化よりも実行時間が短くなったことがわかります (約 0.2 時間)。
fitcauto によって返される最終的なモデルが、観測された最適な学習器となります。モデルを返す前に、関数は学習データ セット全体 (adultdata)、リストされている Learner (またはモデル) のタイプ、および表示されたハイパーパラメーター値を使用して、モデルの再学習を行います。
テスト セットのパフォーマンスの評価
混同行列と受信者動作特性 (ROC) 曲線を使用して、返されたモデル bayesianMdl と ashaMdl の性能をテスト セット adulttest で評価します。
各モデルについて、テスト セットの予測されるラベルとスコア値を見つけます。
[bayesianLabels,bayesianScores] = predict(bayesianMdl,adulttest); [ashaLabels,ashaScores] = predict(ashaMdl,adulttest);
テスト セットの結果から、混同行列を作成します。対角要素は、特定のクラスの正しく分類されたインスタンスの数を示しています。非対角要素は誤分類した観測値のインスタンスです。1 行 2 列のタイル レイアウトを使用して結果を比較します。
tiledlayout(1,2) nexttile confusionchart(adulttest.salary,bayesianLabels) title("Bayesian Optimization") nexttile confusionchart(adulttest.salary,ashaLabels) title("ASHA Optimization")
各モデルについて、テスト セットの分類精度を計算します。精度は、テスト セットの正しく分類された観測値の割合です。
bayesianAccuracy = (1-loss(bayesianMdl,adulttest,"salary"))*100bayesianAccuracy = 85.3243
ashaAccuracy = (1-loss(ashaMdl,adulttest,"salary"))*100ashaAccuracy = 85.4724
混同行列と精度の値から、どちらのモデルも十分な性能を示していることがわかります。
それぞれのモデルについて、ROC 曲線をプロットし、ROC 曲線の下の領域 (AUC) を計算します。ROC 曲線は、分類スコアのさまざまなしきい値についての真陽性率と偽陽性率の関係を示します。しきい値にかかわらず真陽性率が常に 1 の完璧な分類器では、AUC = 1 になります。観測値をランダムにクラスに割り当てるバイナリ分類器では、AUC = 0.5 になります。大きな AUC 値 (1 に近い) は、分類器の性能が高いことを示します。
それぞれのモデルについて、rocmetrics オブジェクトを作成し、ROC 曲線のメトリクスを計算して AUC の値を求めます。
bayesianROC = rocmetrics(adulttest.salary,bayesianScores,bayesianMdl.ClassNames); ashaROC = rocmetrics(adulttest.salary,ashaScores,ashaMdl.ClassNames);
rocmetrics の関数 plot を使用して、ラベル <=50K の ROC 曲線をプロットします。
figure [r1,g1] = plot(bayesianROC,"ClassNames","<=50K"); hold on [r2,g2] = plot(ashaROC,"ClassNames","<=50K"); r1.DisplayName = replace(r1.DisplayName,"<=50K","Bayesian Optimization"); r2.DisplayName = replace(r2.DisplayName,"<=50K","ASHA Optimization"); g1(1).DisplayName = "Bayesian Optimization Model Operating Point"; g2(1).DisplayName = "ASHA Optimization Model Operating Point"; title("ROC Curves for Class <=50K") hold off
AUC 値を基準にすると、どちらの分類器もテスト データに対して適切に機能しています。
参考
fitcauto | confusionchart | perfcurve | BayesianOptimization
関連するトピック
You can also select a web site from the following list:
Americas
- América Latina (Español)
- Canada (English)
- United States (English)
Europe
- Belgium (English)
- Denmark (English)
- Deutschland (Deutsch)
- España (Español)
- Finland (English)
- France (Français)
- Ireland (English)
- Italia (Italiano)
- Luxembourg (English)
- Netherlands (English)
- Norway (English)
- Österreich (Deutsch)
- Portugal (English)
- Sweden (English)
- Switzerland
- United Kingdom (English)