Nearest Neighbor Based Feature Selection for Regression and Its Application to Neural Activity
Published in Advances in Neural Information Processing Systems (NIPS), 2006
Proposed a k-nearest neighbor-based feature selection method for regression, improving predictive accuracy in neural activity analysis. Applied to motor cortex data, the method identified significant features influencing hand velocity prediction.
Recommended citation: Navot et al (2006). "Nearest Neighbor Based Feature Selection for Regression and Its Application to Neural Activity." Advances in Neural Information Processing Systems.
Download Paper