Computational Models for Intelligence and Automation
Machine Learning Algorithms Performance on BCI Competition 4 Dataset 4
Authors: Gauttam Jangir, Gaurav Purohit and Nisheeth Joshi
Publishing Date: 28-08-2025
ISBN: 978-81-975670-1-8
Abstract
Brain computer interface (BCI) is evolving rapidly, as machine learning algorithms are getting better. Algorithms performance plays central role while selecting for real time use cases. One such case is robotic fingers movement, as they play important role in BCI due to their involvement in day-to-day task. In this paper one such standard dataset “BCI competition 4 dataset iv” is considered to evaluate the performance of various machine learning algorithms like Linear Regression (LR),Bayesian Ridge (BR), Light Gradient Boosting (LGB) and decision Tree (DT). Algorithms performance also evaluated on pre-processed dataset. The final result is presented with highest achieved correlation value as 0.76 (noisy dataset) and 0.80 (cleaned dataset) by using the DT.
Keywords
Brain computer interface(BCI), Algorithms, robotic fingers, performance, Linear Regression (LR), Bayesian Ridge (BR), Light Gradient Boosting (LGB) and decision Tree (DT).
Cite as
Gauttam Jangir, Gaurav Purohit and Nisheeth Joshi, "Machine Learning Algorithms Performance on BCI Competition 4 Dataset 4", In: Puneet Kumar Gupta (eds), Computational Models for Intelligence and Automation, SCRS, India, 2025, pp. 70-84. https://doi.org/10.56155/978-81-975670-1-8-7