How to decide value of 'ndim' when using ,[residuals,reconstructed] = pcares(X,ndim) , for Feature size reduction using PCA?
1 Ansicht (letzte 30 Tage)
Ältere Kommentare anzeigen
I am working on feature classification using KNN and SVM. Data size is 2000 images and features are histograms (of various bin sizes ) of large size for each image. I read above function can be used but my query is how to decide value of ndim which will ensure that the best features are retained ?
0 Kommentare
Antworten (0)
Siehe auch
Kategorien
Mehr zu Dimensionality Reduction and Feature Extraction finden Sie in Help Center und File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!