NeuralNet net;
net.firstLayer(15,"dataset.csv").addLayer(5).lastLayer("target.csv");
net.train();
net.infere("test.csv");
- dataset.csv: contains the data in the form of one N-dimensional point per row
- target.csv: contains the class (+1 or -1) of points in dataset
- test.csv: has some points to test the data
NeuralAnalyzer is able to generate dataset, test file and to plot all the data.
You can compile it in root running
.x compile.C
An example is in run_plot.cxx
XRayMachine plots the architecture design of the net simply Example of usage is:
auto cx = XRayMachine(net).fCanvas;
An example is in run_xray.cxx