A system for behavior prediction based on neural signals

Authors:
Jacob Mathew;Laxmikanta Sahoo;Goutam Saha
Affiliations:
Department of Electronics and Electrical Communication Engineering, IIT Kharagpur, Kharagpur, West Bengal, 721302, India;Department of Electronics and Electrical Communication Engineering, IIT Kharagpur, Kharagpur, West Bengal, 721302, India;Department of Electronics and Electrical Communication Engineering, IIT Kharagpur, Kharagpur, West Bengal, 721302, India
Venue:
Neurocomputing
Year:
2012

Citing 10
Cited 0

Original Contribution: A scaled conjugate gradient algorithm for fast supervised learning

Neural Networks
Machine Learning

Machine Learning
Pattern Classification (2nd Edition)

Pattern Classification (2nd Edition)
Unsupervised spike detection and sorting with wavelets and superparamagnetic clustering

Neural Computation
Pattern Recognition and Machine Learning (Information Science and Statistics)

Pattern Recognition and Machine Learning (Information Science and Statistics)
Sorting of neural spikes: When wavelet based methods outperform principal component analysis

Natural Computing: an international journal
Neural signal classification using a simplified feature set with nonparametric clustering

Neurocomputing
Letters: Solving alignment problems in neural spike sorting using frequency domain PCA

Neurocomputing
Decision functions that can support a hippocampal model

Neurocomputing
Maximum subset intersection

Information Processing Letters

Quantified Score

Hi-index	0.01

Visualization

Abstract

Modeling the brain has been a major scientific endeavor in the last few decades. Models can be at different levels of abstraction like chemical, cellular, region etc. Limitations in physical experimental setup will create poor model, if a regular pattern classifier system is used. This paper proposes a pattern recognition system to classify/predict animal behavior based on hippocampal signals. This system, takes into account the practical constraints during acquisition of the neural signal. The system performance for action classification is tested using multiple databases and percentage misclassification as low as 3.25% was obtained.