photo

Greg Heath


Last seen: 約4年 前 |  2011 年からアクティブ

Followers: 0   Following: 0

メッセージ

Backgound in Electromagnetic Theory, Plasma Physics and Radar Target Identification using Neural Networks.
PhD Student, Research Assistant and Lecturer at Stanford;
AB, ScB, ScM Student; Research Assistant, Fellow and Professor at Brown;
27 yrs researching Ballistic and Theatre Missile Defense using Neural Networks at MIT Lincoln Laboratory. Retired 2003.

PLEASE DO NOT SEND QUESTIONS AND DATA TO MY EMAIL. HOWEVER, CAN SEND LINKS TO POSTS.
Professional Interests: Neural Netwoks, Spectral Analysis

統計

  • First Review
  • 36 Month Streak
  • Thankful Level 5
  • Ace
  • Revival Level 4
  • Knowledgeable Level 5
  • First Answer

バッジを表示

Feeds

表示方法

回答済み
Would this be considered underfitting?
A model is UNDERFIT if and only if No. of independent training equations < No. of unknowns Hope this helps Tha...

約4年 前 | 0

回答済み
Avoid exploding/vanishing gradient problem with NARX nets?
Use a higher open-loop peformance goal. Then lower the value after the loop is closed. It's been years since I've done this so ...

約4年 前 | 0

回答済み
How to make prediction from a trained NARX neural network?
You forgot to include the intial conditions: yz = nets(xz,xiz,aiz); Thank you for formally accepting my answer Greg

4年以上 前 | 0

回答済み
Out of memory issue while training a Neural Network (NN), array exceeds maximum array size preference using backpropJacobianStatic
A single hidden layer is sufficient. Hope this helps Thank you for formally accepting my answer Greg

4年以上 前 | 0

回答済み
Out of memory issue while training a Neural Network (NN), array exceeds maximum array size preference using backpropJacobianStatic
A single hidden layer is sufficient. Hope this helps Thank you for formally accepting my answer Greg

4年以上 前 | 0

回答済み
Difference between function of sim and predict in Neural network?
help sim doc sim help predict doc predict Thank you for formally accepting my answer Greg

4年以上 前 | 0

回答済み
Forecasting using GRNN (Generalized Regression Neural Network)?
Yes, Normalize input and output data to [-1,1]. Minimize the order for a sufficient polynomial fit. Minimize the number of ...

4年以上 前 | 0

回答済み
CNN With unbalanced Data
Although you do not prefer data augmentation, duplication of the smaller dataset examples is probably the quickest and most reli...

4年以上 前 | 0

質問


How to find questions on neural neural networks.
I have been away for awhile: How can I list the neural network questions? Greg

4年以上 前 | 1 件の回答 | 0

1

回答

回答済み
Trained neural network to optimize input variable
"OPTIMIZED INPUT" is ill-defined Just train another net x2 = f2(x1,y1,y2) Hope this helps, THANK YOU FOR FORMALLY ACCEP...

4年以上 前 | 0

回答済み
Linear Integer Output from a Neural Network
The integer nature of the output DOES NOT HAVE TO BE CONSIDERED during training. It is sufficient to round the real valued ou...

4年以上 前 | 0

回答済み
How to disable validation and test data set in neural network
You have to define net before modifying any properties. clear all, close all, clc [x,t] = iris_dataset; for i = 1:2 net ...

5年弱 前 | 1

回答済み
Training data and Training target in Neural Networks
You cannot make any intelligent decisions until you have examined a plot of the data!!! (WRONG!!! Plotting the data first is ...

5年弱 前 | 0

| 採用済み

回答済み
simulink neural network producing different outputs to workspace
A simpler solution is to ALWAYS begin the program with a resetting of the random number generator. For example, choose your favo...

5年弱 前 | 1

回答済み
How to avoid getting negative values when training a neural network?
Use a sigmoid for the output layer. Hope this helps THANK YOU FOR FORMALLY ACCEPTING MY ANSWER GREG

5年弱 前 | 0

回答済み
weird plotregression plots for a 10% of my fitnet neural networks
Sometimes training gets into a parameter space rut. That is why it is wise to train multiple models. Hope this helps. Greg ...

5年弱 前 | 0

回答済み
What is the purpose of shuffling the validation set?
To impose and verify a consistent GENERALIZED path to convergence by avoiding repetitive anomalies. Hope this helps Greg

5年弱 前 | 0

回答済み
How to force overfiting of Deep Learning Network for Classification
OVERFITTING = More training unknowns (e.g., weights) than training vectors. OVERTRAINING1 = Training an overfit network to...

5年弱 前 | 0

回答済み
Neural Network Pattern Recognition
Targets are 1-dimensional unit vectors with 4 zeros and a single 1 . Thank you for formally accepting my answer. Greg

5年弱 前 | 0

| 採用済み

回答済み
semanticseg producing marginally different values when inference is repeated
Use clear all, close all, clc, rng(0) on the 1st line Thank you for formally accepting my answer Greg

5年弱 前 | 0

回答済み
combining two neural networks (net1 is trained & net2 is untrained) in one bigger network
Just 1. Save the outputs of net1 in a file 2. Use the file to train net2 Greg

5年弱 前 | 0

回答済み
Problem with the TreeBagger Command
The sizes of the input function and output target must be [ I N ] = size(input) [ O N ] = size (target) Hope this helps...

5年弱 前 | 0

回答済み
How do I create a neural network that will give multiple input and outputs?
ALWAYS arrange your data so that [ I N ] = size(input) [ O N ] = size(output) Hope this helps. Greg

5年弱 前 | 0

回答済み
How to make a hybrid model (LSTM and Ensemble) in MATLAB
Replace your YES/NO data with either 1/0 or 1/-1. Hope this helps. Greg

5年弱 前 | 1

回答済み
how to augment image data only for a specific class?
Separate class 0 and interpolate. If you have a good feel for the data you could extrapolate. However the latter might be tricky...

5年弱 前 | 0

回答済み
NTSTOOL - How to get predicted values of "the future"?
The known time series is analyzed to yield a time-series model that uses past and present values to predict future values. The ...

約5年 前 | 0

回答済み
Timedelaynet output calculation principle
You did not include tHe 2 biases. Hope this helps. Greg THANK YOU FOR FORMALLY ACCEPTING MY ANSWER

約5年 前 | 0

回答済み
Understand number of weights of Neural Network
It is possible. In general, however, you don't have the slightest idea what choice would be significantly better than random. ...

約5年 前 | 0

回答済み
Using pca for features selections
PCA (Principal Coordinate Analysis) is a very useful method for regression (it ranks linear combinations of the original variabl...

約5年 前 | 0

回答済み
How do we decide the number of hiddenlayers in a PatternNet?
patternnet(10) indicates ONE HIDDEN LAYER WITH TEN NODES It is important to be mindful of the number of layers and nodes. The ...

5年以上 前 | 0

| 採用済み

さらに読み込む