- The abstract indicates "The framework combines a pre-trained convolutional neural network, OpenL3, and a support vector machine (SVM) classifier."
- The OpenL3 network is easily usable in MATLAB https://www.mathworks.com/help/audio/ref/openl3.html
- The methods section indicates "This experiment is implemented based on MATLAB R2022a."
- The conclusions section indicates "Through testing on the VOICED dataset, the proposed method achieves 99.46%, 99.64%, 98.92%, and 99.64% values for the ACC, SEN, SPE, and F1 metrics, respectively. Compared with the existing works and the compared machine learning methods, the proposed method exhibits better performance."
How can I increase classifier accuracy?
2 ビュー (過去 30 日間)
古いコメントを表示
I'm working on matlab code for detecting speech disorder, I've five classes each class is 220 sounds, so I use calssification learner in matlab with some features(statistic features, spectrm and MFCCs), The problem is that the accuracy of classifier is weak (about 75%), How can I increase it?
0 件のコメント
回答 (1 件)
Drew
2023 年 6 月 8 日
編集済み: Drew
2023 年 6 月 8 日
Voice disorder classification is an active area of research. There are many papers on the topic. The paper "Voice disorder classification using convolutional neural network based on deep transfer learning", published in May 2023, is one example (see https://www.nature.com/articles/s41598-023-34461-9). Some observations:
0 件のコメント
参考
カテゴリ
Help Center および File Exchange で Classification についてさらに検索
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!