mlperceptron 0.3

Multilayer Perceptron in Python

Python implementation of multilayer perceptron neural network from scratch.

Minimal neural network class with regularization using scipy minimize. Contains clear pydoc for learners to better understand each stage in the neural network.

https://github.com/paulokuong/mlperceptron

Requirements

• Python 3.4 (tested)

Goal

Code sample

X = 000, 001, 010, 011, 100, 101, 110, 111

y = 0,0,0,0,0,0,1

Data we want to predict: p = 011, 111, 000, 010, 111 Expected results are: 0, 1, 0, 0, 1

import numpy as np
from mlperceptron.mlperceptron import NeuralNetwork

X = np.matrix(
    '0 0 0;0 0 1;0 1 0;0 1 1;1 0 0;1 0 1;1 1 0;1 1 1')
y = np.matrix('0;0;0;0;0;0;0;1')
n = NeuralNetwork((5,5,))

g = n.train(X, y, 0.01, show_cost=True)
y_pred = n.predict(np.matrix('0 1 1;1 1 1;0 0 0;0 1 0;1 1 1'), g)

print(y_pred)
print(n.accuracy(y_pred, np.matrix('0;1;0;0;1')))

Contributors

• Paulo Kuong (@pkuong)