Assignment 2 Machine Learning and Data Mining - Handwriting Recognition

Project for assignment 1 of Machine Learning and Data Mining course

This project has 5 main files:

Handwriting.py - Contains the main class for prediction, it also has embededd load file and cross validation features
Main.py - Main file/class that can depending on the parameters passed
- Run all experiments ( 8 algorithms with different Parameters ) in multiprocessing way
- Run all experiments in sequential way
- Run the two best Algorithms with 3 PCA variations
- Display the first 100 samples of the dataset as an image (16x16 Matrix)
results.log: this is the file generated after executing the code with meaningful information about the execution timeline
report: folder with latex files to create the report
semeion.data: file contained 1593 samples of hadwriting digits with 256 features each corresponding to a 16x16 matrix

How to run the code

In order to run the code you go to the root folder and run

python Main.py

This will run the default mode of the project. Which is running the 2 best algorithms with 3 variations of PCA

If you want to use other modes you can run the project with parameters:

python Main.py [all_multi | all_seq | best | image]

This will run the algorithm in the one of the modes explained previously