SIGNATURE WORK
CONFERENCE & EXHIBITION 2024

This project demonstrates the value of integrating machine learning techniques in League of Legends for strategic decision-making and accurate predictions. By utilizing the Riot API, LCU, and machine learning algorithms, pre-game and in-game prediction applications were developed. Data from the Riot API and statistics websites were gathered, and word embedding techniques were applied to provide insights on champion and item compatibility. LSTM models were used for real-time in-game win rate prediction, particularly important in high-stakes tournaments. The project includes a Tauri application for pre-game prediction, empowering players to make informed decisions and enhance their performance. Additionally, a frontend overlay for in-game prediction adds value to the viewing experience of LoL matches in spectating mode. It showcases the growing significance of machine learning in League of Legends and highlights the potential for further development using Riot API, LCU API, and machine learning algorithms.