Probabilistic modeling of operant conditioning behaviors via stochastic DNA strand displacement cascades

Operant conditioning is a learning mechanism by which animals adapt to its external environment and past experiences. In the field of artificial intelligence, DNA strand displacement (DSD) technology has performed well in various aspects. Chemical reaction networks (CRNs) are constructed using stochastic DSD technology to study operant conditioning, and the simulation results are verified by Visual DSD software. In this paper, the DSD technology is utilized to construct CRNs to achieve different kinds of of learning and forgetting processes and generalization in operant conditioning. A comparative analysis is carried out on the four simulation results, and the peak acquisition values of each experiment are compared. The stochastic DSD technology is used to design stochastic CRNs to construct probabilistic decision making systems. The two-way probabilistic decision making of and the three-way probabilistic decision making of animal behaviors are studied. This paper presents the weight variations for each experiment in tabular form. Finally, a comparative analysis is conducted on the probabilistic outcomes of the two-way and three-way probabilistic decision-making experiments. CRNs can be used to achieve realistic behaviors in engineered bionic systems. It provides a direction for the integration of biology and psychology.