readme.md

Using machine learning for predicting League of Legends match outcome

This repository contains the implementation of related article that was written as a part of Machine learning course at the Faculty of Computer and Information Science at the University of Ljubljana.

Repository structure

The repository contains the following folders:

• article contains the article
• classifiers contains implementations of the solutions for the specified problem
• figs contains images of plots
• old contains old python programs, that were not used in the final implementations
• prepare_data.py python program for processing the raw data
• time_comparison.ipynb Jupyter Notebook file for comparing classifiers by their learning time and accuracy

Data

The data set that was used for buliding and testing the models can be accessed here.

Reproducing results

To run the Python programs and reproduce the results I got, you will have to download the Kaggle data set and extract it to folder \data\raw_data.

After you successfully forked this repository and downloaded the data set from Kaggle you will have to install python dependencies listed in requirements.txt. You can do this by using pip:

pip install -r requirements.txt