Neural encoding based on frequency states using multi-spike train data

Understanding the way how the neural networks encode outer stimulus and inner decision-making process is a key problem in neuroscience as well as in artificial intelligence. Although researchers have proposed several assumptions and related models, the supporting biological evidences are rarely provided. Our task involves finding an encoding method based on neuron firing frequency states and its transformation model under certain stimulus. Besides, a new method to analyze spikes train data is also proposed which is proved effective here. Then, the results by analyzing multi-microelectrodes simultaneous recorded spike train from temporal cortex and hippocampus of mouse experiment is shown, supporting and validating this model.