Error Correcting Output Codes (ECOC) Classifier
Error Correcting Output Codes (ECOC) for multi-class classification.
現在この提出コンテンツをフォロー中です。
- フォローしているコンテンツ フィードに更新が表示されます。
- コミュニケーション基本設定に応じて電子メールを受け取ることができます
"Error Correcting Output Codes (ECOC) for multi-class classification".
Different code matrices i.e. one vs.one, one vs.All, dense random and sparse random implemented. You have also opportunity to cheese between MLP neural network and Support Vector Machine Classifiers.
- Open Matlab, change directory and Run "Demo.m"
- In demo, I use Segment dataset just to show how the code works. You should have your dataset prepared in that format.
- To read more, go to folder "papers".
Citations:
- Nima Hatami: Thinned-ECOC ensemble based on sequential code shrinking. Expert Syst. Appl. 39(1): 936-947 (2012)
- Giuliano Armano, Camelia Chira, Nima Hatami: Ensemble of Binary Learners for Reliable Text Categorization with a Reject Option. HAIS (1) 2012: 137-146
- Giuliano Armano, Camelia Chira, Nima Hatami: Error-Correcting Output Codes for Multi-Label Text Categorization. IIR 2012: 26-37
引用
Nima Hatami (2026). Error Correcting Output Codes (ECOC) Classifier (https://jp.mathworks.com/matlabcentral/fileexchange/47405-error-correcting-output-codes-ecoc-classifier), MATLAB Central File Exchange. に取得済み.
