MLP Class Reference

Multilayer Perceptron class. Performs binary and categorical classification Uses nn::Linear as a sequential model, follows PyTorch's implementation. TODO: Addition of Conv1d layers. More...

#include <mlp.h>

Public Member Functions
	MLP (std::vector< std::vector< double > > const &, std::vector< std::pair< int, int > > const, const int epochs=100, const double learning_rate=0.001)
	default constructor for MLP class
void	fit ()
	fit an MLP on the input data
double	predict (std::vector< double > const &)
	performs inference

Detailed Description

Multilayer Perceptron class. Performs binary and categorical classification Uses nn::Linear as a sequential model, follows PyTorch's implementation. TODO: Addition of Conv1d layers.

Constructor & Destructor Documentation

MLP()

MLP::MLP ( std::vector< std::vector< double > > const & data, std::vector< std::pair< int, int > > const arch, const int epochs = 100, const double learning_rate = 0.001 )

inlineexplicit

default constructor for MLP class

Parameters

data	2D vector, The input data. As usual, the last element of each sub-vector represents the label of the row
arch	1D vector of pairs. Represents the [in_features, out_features] of each layer in the network
epochs(int)	The number of epochs
learning_rate(double)	The learning rate

Member Function Documentation

predict()

double MLP::predict ( std::vector< double > const & input )

inline

performs inference

Parameters

input

1D vector, the passed validation data

Returns

double: The classified label

---

The documentation for this class was generated from the following file:

• /Users/runner/work/AlgoPlus/AlgoPlus/src/machine_learning/nn/mlp.h