The brain´s neural classifiers considering both the posterior probabilities and generalities to control the mechanism underlying decision making; an evidence for computational Bayesian classifiers

To study a cognitive neural model of decision making, we analyzed the neural and behavioral data recorded in Shadlen Neuroscience Lab [9] from the monkeys performing motion-discrimination reaction-time task with consideration of six coherence levels. Two uncorrelated principal components of the timing sequences of each trial´s action potential have been further extracted to examine the existing information from the spike trains. The trials corresponding to right and wrong choices were analyzed independently to determine whether the Bays´ rule can describe the decision making mechanism or not. The result demonstrates that the brain generates the spikes to temporally extract two principal components leading to making a decision: posterior probabilities and generalities. At the end, the temporal model of Bayesian decision making has been theoretically described and verified through examination of above data.