smmlp

About

It's a simple multilayer perceptron model, implemented in c++ to get an inner idea about backpropagation.

Backpropagation

More about this on brilliant.org

File Description

For a simple MLP model i wrote 6 important file:

• layer.cpp
• node.cpp
• network.cpp
• sample.h
• utils.h
• main.cpp

main.cpp: The smmlp models entry point where you can define your model structure and process traning and test data for your model.

layer.cpp: With layer.h where i define all layers update/matrix manipulation.

node.cpp: with node.h where i define node's update/matrix manipulation

network.cpp: with network.h where i define network structure (I mean number of neuron, number of layer)

sample.h: Training sample structure

utls.h: Different kind of helper function

Note

Don't forget to say thanks davidalbertonogueira. Most of the code are borrowed from him.

My blog: smafjal