a standard LeNet2/LeNet5 architecture without the last RBF layer (look
into net-cscscfe for that).
8.1.0.7.12.0. (new net-csccf iniinjki0kj0tbl0si0sj0ki1kj1tbl1ki2kj2tbl2outthickprm)
[CLASS] (packages/gblearn2/net-csccf.lsh)
makes a new net-csccf module. ini
inj : expected max size of input for preallocation of
internal states ki0
kj0 : kernel size for first convolutional layer
tbl0 : table of connections between input anf feature maps
for first layer si0
sj0 : subsampling for first layer ki1
kj1 tbl1 : kernel and table
for next layer ki2
kj2 tbl2 : same for last
convolution layer outthick : number of
outputs. prm an idx1-ddparam in which
the parameters will be allocated.