Machine Learning Approaches to Choose Heroes in Dota 2

• Published in: Proceedings of the XXth Conference of Open Innovations Association FRUCT pp. 345 - 350
• Main Authors: Porokhnenko, Iuliia, Polezhaev, Petr, Shukhman, Alexander
• Format: Conference Proceeding
• Language: English
• Published: FRUCT 01.04.2019
• Subjects: Classification algorithms; Games; Loss measurement; Machine learning; Machine learning algorithms; Training; Vegetation
• ISSN: 2305-7254
• DOI: 10.23919/FRUCT.2019.8711985

The winning in the multiplayer online game Dota 2 for teams is a sum of many factors. One of the most significant of them is the right choice of heroes for the team. It is possible to predict a match result based on the chosen heroes for both teams. This paper considers different approaches to predicting results of a match using machine learning methods to solve the classification problem. The experimental comparison of predictive classification models was done, including the optimization of their hyperparameters. It showed that the best classification models are linear regression, linear support vector machine, as well as neural network with Softplus and Sigmoid activation functions. The fastest of them is the linear regression model, so it is best suited for practical implementation.