mlp.cpp

Example on using the Multi-Layer Perceptrons networks supplied from pygpmp

Todo:

: TODO

 
#include <deque>
#include <fstream>
#include <iostream>
#include <openGPMP/ml.hpp>
 
// use the openGPMP Machine Learning namespaces
using namespace gpmp::ml;
 
int main() {
    std::cout << "SAMPLE USING THE MULTI-LAYER PERCEPTRON NETWORK.\n";
 
    std::srand(42069);
    // initialize the SecondaryMLP model
    int64_t inputs{1}, outputs{1}, hidden_layer_units{8}, hidden_layers{3};
 
    long double lr{.5};
    gpmp::ml::Trainers util;
    auto model = util.init_SecondaryMLP(inputs = 1,
                                        outputs = 1,
                                        hidden_layer_units = 8,
                                        hidden_layers = 3,
                                        lr = .5);
 
    // declare a data file
 
    // TODO : create util method to save/read/parse datafile, and
    // columns
    // std::ofstream data_file;
    std::string data_file = "../../data/school_scores.csv";
 
    uint64_t max_iters{1000}, return_every{500};
    // call the train method for the SecondaryMLP model
    util.train_SecondaryMLP(model, data_file, max_iters, return_every);
 
    return 0;
}