fenixnix / BP-ANN

Ann BP algorithm impliment in Cpp&Lua

BP-ANN

BP ANN Frameware by Nix

NNerve

A nerve cell You can use it to build an perceptron to solve AND OR problem but not XOR

NLayer

A ANN layer consist of NNerve

NAnnBp

A BP-ANN consist of NLayer You can use it to build an multilayer ANN to solve some classic problem such as classify

#Easy to Use!!!

    NAnnBp bp;

//Init Bp Ann with 2Node InputLayer 5Node HideLayer1 3Node HideLayer2 and 1Node OutputLayer
    vector<int> layerNums = {2,5,3,1};
    bp.Init(layerNums);

//Learn And 
    vector<float> inputs = {0,0};
    vector<float> outputs = {0};
    bp.Learn(inputs,outputs);

    vector<float> inputs = {1,0};
    vector<float> outputs = {0};
    bp.Learn(inputs,outputs);

    vector<float> inputs = {0,1};
    vector<float> outputs = {0};
    bp.Learn(inputs,outputs);

    vector<float> inputs = {1,1};
    vector<float> outputs = {1};
    bp.Learn(inputs,outputs);

//Ask and Get Result
    vector<float> inputs = {0,1};
    auto res = bp.Run(inputs);

    cout<<"\tO:";
    foreach(auto r , res){
        cout<<r<<",";
    }
    cout<<endl;