Description: Independent Learning the encoder and the sparse classifiers achieve the combination. First through sparse coding since no label was handwritten 5-9 training obtain the optimal parameters, and then through the front propagation, get the training and test sets of features, a label by 0-4 trained softmax model train set, then enter the test set to the classification model to classify.
To Search:
File list (Check if you may need any files):
self-taught learning
....................\display_network.m
....................\feedForwardAutoencoder.m
....................\initializeParameters.m
....................\loadMNISTImages.m
....................\loadMNISTLabels.m
....................\mnist-train-images.idx3-ubyte
....................\mnist-train-labels.idx1-ubyte
....................\softmaxCost.m
....................\softmaxPredict.asv
....................\softmaxPredict.m
....................\softmaxTrain.m
....................\sparseAutoencoderCost.m
....................\stlExercise.asv
....................\stlExercise.m
Download