@inproceedings{cf53a20d88234d20981613e9f0fa09e5,
title = "Efficient support vector regression with reduced training data",
abstract = "Support Vector Regression (SVR) as a supervised machine learning algorithm have gained popularity in various fields. However, the quadratic complexity of the SVR in the number of training examples prevents it from many practical applications with large training datasets. This paper aims to explore efficient ways that maximize prediction accuracy of the SVR at the minimum number of training examples. For this purpose, a clustered greedy strategy and a Genetic Algorithm (GA) based approach are proposed for optimal subset selection. The performance of the developed methods has been illustrated in the context of Clash Royale Challenge 2019, concerned with decks' win rate prediction. The training dataset with 100,000 examples were reduced to hundreds, which were fed to SVR training to maximize model prediction performance measured in validation R2 score. Our approach achieved the second highest score among over hundred participating teams in this challenge.",
keywords = "Clash Royale, Greedy search, K-means clustering, R-squared metric, Support Vector Regression (SVR)",
author = "Ling Cen and Vu, {Quang Hieu} and Dymitr Ruta",
note = "Publisher Copyright: {\textcopyright} 2019 Polish Information Processing Society - as since.; 2019 Federated Conference on Computer Science and Information Systems, FedCSIS 2019 ; Conference date: 01-09-2019 Through 04-09-2019",
year = "2019",
month = sep,
doi = "10.15439/2019F362",
language = "British English",
series = "Proceedings of the 2019 Federated Conference on Computer Science and Information Systems, FedCSIS 2019",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "15--18",
editor = "Maria Ganzha and Leszek Maciaszek and Leszek Maciaszek and Marcin Paprzycki",
booktitle = "Proceedings of the 2019 Federated Conference on Computer Science and Information Systems, FedCSIS 2019",
address = "United States",
}