Digit Classification with Kernel Perceptron - Author: Katia Cracco
Project developed for the Statistical Methods for Machine Learning Course in Università degli Studi di Milano
preprocessing.py: this file pre-processes the dataset present in the dataset folder (zipped) by generating the files present in the input folder (zipped); it also generates polynomial kernels which have not been loaded because they are too heavy.
digit.py: it contains the main of the principal program, that is of the algorithm to be implemented.
multiClassKernelPerceptron.py: this class contains the methods called by digit.py
kernelPerceptron.py: this class contains the methods called by multiClassKernelPerceptron.py