Author(s)

Xiaolin Sun, Baolin Liu

Corresponding Author

Baolin Liu

ABSTRACT

Humans can identify actions performed by others rapidly and accurately. Here we applied machine learning techniques in functional magnetic resonance imaging (fMRI) data to reveal the neural mechanism underlying such neural activity. We used searchlight multi-voxel pattern analysis method based on support vector machine classifier to identify the regions that can discriminate different actions significantly, so as to clarify the brain regions recruited for action identifications. Our results indicated that part of the regions in occipito-temporal and frontal cortex hold the ability to identify different human actions.

KEYWORDS

Machine learning, Support vector machine, Functional magnetic resonance imaging, Multi-voxel pattern analysis