void train(int x, int j)
{
for (int i = 0; i < fea_num; ++i)
mlp[0][i] = neuron(0, i, 0, fea_train[x][i]);
for (int i = 0; i < tag_num; ++i)
mlp[0][i+fea_num] = neuron(0, fea_num+i, 0, tag_train[x][i]);
mlp[0][fea_num+tag_num] = neuron(0, fea_num+tag_num, 0, 1);
//for (int i = 0; i < 10; ++i)
// ans[i] = concept[i][id_train[x]];
ans[0] = concept[j][id_train[x]];
ans[1]=1-ans[0];
calc();
adjust();
if (concept[j][id_train[x]]==1)
{
if (mlp[layers-1][0].y>=mlp[layers-1][1].y)
++pgood;
++psum;
}
else
{
if (mlp[layers-1][0].y<=mlp[layers-1][1].y)
++ngood;
++nsum;
}
}
void test(int x, int k)
{
for (int i = 0; i < fea_num; ++i)
mlp[0][i] = neuron(0, i, 0, (k==1?fea_test:fea_database)[x][i]);
for (int i = 0; i < tag_num; ++i)
mlp[0][i+fea_num] = neuron(0, fea_num+i, 0, (k==1?tag_test:tag_database)[x][i]);
mlp[0][fea_num+tag_num] = neuron(0, fea_num+tag_num, 0, 1);
calc();
}
void NetworkLayer :: init(int nNeurons){
N = nNeurons;
for(int i =0;i<nNeurons;i++){
Neuron neuron(nNeurons);
// printf("Generating neurons on layer %d , neuron id is %d\n", layerId, neuron.getID());
Neurons.push_back(neuron);
}
}
int main()
{
srand(0);//(unsigned)time(NULL));
initialize();
for (int k = 0; k < 10; ++k)
{
for (int i = 1; i < layers; ++i)
for (int j = 0; j < neu_num[i]; ++j)
mlp[i][j] = neuron((i < layers - 1 ? inner : outer), j, i, 0);
train(k);
test(k);
}
output();
return 0;
}
