We carry out theoretical research on life and ecosystem, especially on computational neuroscience. Based on experimental observation, we establish neural networks with specific function to study cognition and self-organization behavior in learning process from the perspective of complexity. We explore the formation mechanism of brain circuits, analyze the emergence and self-organization in learning process, and give advice on experimental design of brain mechanism.

Economy, ecology and environmental coupling evolution

Metabolism is a distinguished feature of organism. Each species in the ecological system are interrelated through the metabolic cycle of energy flow and form a unified network. According to metabolic ecology and ecological neutral theory, the distribution of these energy flows in biological individual and ecological community is of obvious statistical characteristics. Our study takes food web as an energy transport network and adopts complex network and Markov chain model to study the macro characteristics of his network. At the same time, we have also been attempting to generalize the allometric scaling of ecosystem, neutral theory of species diversity into complex system, such as food web, city, and nation.

Learning and brain dynamics

Learning is a process form being unable to being able, or the improvement and transference in behavior. It is the transition of cognitive function status. From the perspective of neuron, learning includes the change of neuron activity and interaction, such as the change of neuron suitability and synaptic strength (e.g. STDP, Hebian plasticity) or the change of neurotransmitter release level (e.g. DA), etc. Based on the nerve electrophysiology and the data of psychophysical experiment, we are attempting to construct neural network and studying the dynamic phenomenon in learning process.

Computational neuroscience

Based on the experimental observation data, with mathematical and physical model, we study complex phenomenon and functions in nervous system, such as concussion of nervous system and core mechanism of higher cognitive function, including decision-making, risk perception, memory and even emotion. Computational neuroscience is a crucial part of noetic science, which was come up with by Qian Xuesen. With neural network model, we find that    multiple-space working memory is adjusted by the breadth and intensity of excitable connection. And we find a mechanism of working memory confusion accordingly. With firing rate model, we find that the saddle point structure of nervous system and dynamic system is the foundation of multiple-choice. We also reveal the temporal characteristic that the initial state of system and the dynamics around saddle point co-determine the multiple-decision.