Main Content
2 results
Rprop training for Artificial Neural Networks
class2lab(nn,class)
- CLASS2LAB Convert RNN classes back to labels
computelayer(inputs,Weights,typelayer)
- COMPUTELAYER Compute the output of a NN layer
opt_rprop(nn,train_in,train_out,test_in,test_out)
- OPT_RPROP Optimize a NN using Rprop
class2matrix(in,n_classes)
- CLASS2MATRIX Convert classes to a binary matrix
Demo_Rprop_2(batch)
- DEMO_RPROP_2 All 4 Rprop on MNIST-small dataset
class_stat(inputs,targets,n_class)
- CLASS_STAT Compute various classification statistics
lab2class(labels,nn)
- LAB2CLASS Convert labels to RNN classes
init_nn(size_layer,nn)
- INIT_NN Initialize a NN
shuffledata(matrix,index)
- SHUFFLEDATA Shuffle the rows of a matrix
computenetwork(nn,inputs,fullout)
- COMPUTENETWORK Compute the output of a NN
indent(verbose)
- INDENT Indent text
Demo_Rprop_1(batch)
- DEMO_RPROP_1 IRprop- on MNIST-small dataset
MSE(inputs,targets)
- MSE Compute the Mean Square Error (MSE)
Convert N-dimensional cells of K-dimensional matrices to N*K matrices.
Demo_cell2num.m
- DEMO_CELL2NUM Demonstrate some features of the NUM2CELL function
cell2num(input,parameters)
- CELL2NUM Convert a cell into a matrix
linearize(X)
- LINEARIZE Linearize a matrix to a vector