Machine learning for neural decoding

TL;DR

This paper reviews how modern machine learning techniques, especially neural networks and ensemble methods, can significantly enhance neural decoding performance over traditional approaches, aiding neuroscience research and brain-machine interface development.

Contribution

It provides practical guidance, comparisons, and code for applying advanced machine learning algorithms to neural decoding tasks, demonstrating their superiority.

Findings

Neural networks outperform traditional decoding methods.

Ensemble methods improve decoding accuracy.

Modern algorithms enhance understanding of neural information.

Abstract

Despite rapid advances in machine learning tools, the majority of neural decoding approaches still use traditional methods. Modern machine learning tools, which are versatile and easy to use, have the potential to significantly improve decoding performance. This tutorial describes how to effectively apply these algorithms for typical decoding problems. We provide descriptions, best practices, and code for applying common machine learning methods, including neural networks and gradient boosting. We also provide detailed comparisons of the performance of various methods at the task of decoding spiking activity in motor cortex, somatosensory cortex, and hippocampus. Modern methods, particularly neural networks and ensembles, significantly outperform traditional approaches, such as Wiener and Kalman filters. Improving the performance of neural decoding algorithms allows neuroscientists to better understand the information contained in a neural population and can help advance engineering applications such as brain machine interfaces.