Scripts for assignment 1 in the COSC420 course. MLP.py implements a multi-layer perceptron (1 hidden layer) with the option of specifiying activation function (logistic or ReLU), distribution and range to draw initial weights from, as well as several other parameters. MLPUserInterface.py implements a slightly modified version of MLP.py with a text based user interface. Expects a 'in.txt' specifying input data, 'teach.txt' specifying target data, and 'param.txt' specifying number of input units, hidden units, output units, learning rate, momentum constant, and error threshold in the same folder.
Plots.Rmd is an R script for creating the graphs used in the report based on the experiments carried out with MLP.py.