Recent Developing a Novel Parameter Estimation Method for an Agent-Based Model in Immune System Simulation Using History Matching: A Case Study on Influenza a Virus Infection| Chapter 4 | Recent Research Advances in Biology Vol. 10

 Agent-based models (ABM) have been widely employed for immune system simulation because they may provide a natural and flexible description of nonlinear dynamic behaviour of complex systems. However, including experimental data into ABM is critical for obtaining an adequate estimation for the model's main parameters. A systematic strategy for immune system simulation is proposed in this research by combining the ABM and regression method within the context of history matching. During the operation, a novel parameter estimate approach is suggested that incorporates the experiment data for the simulator ABM. To begin, we use ABM as a simulator to model the situation. system of defence Then, utilising the ABM's input and output data, the dimension-reduced type generalised additive model (GAM) is used to train a statistical regression model and serve as an emulator during history matching. Next, we introduce an implausible measure to exclude the implausible input values, reducing the parameter input space. Finally, the particle swarm optimization algorithm (PSO) is used to estimate model parameters by fitting the data. system of defences Then, utilising the ABM's input and output data, the dimension-reduced type generalised additive model (GAM) is used to train a statistical regression model and act as an emulator during history matching. Following that, we provide an implausible measure to eliminate the implausible input values, reducing the parameter input space. Finally, the particle swarm optimization technique (PSO) is used to estimate model parameters by fitting the data to the model.  Among the non-implausible input values, there is data from an experiment. The performance of our suggested method is demonstrated using a genuine Influenza A Virus (IAV) data set, and the results show that the proposed method not only has good fitting and prediction accuracy, but also has good computational efficiency.

Author (s) Details

Le Zhang
College of Computer Science, Sichuan University, Chengdu 610065, China.

Tingting Li
College of Mathematics and Statistics, Southwest University, Chongqing 400715, China.

View Book :-  https://stm.bookpi.org/RRAB-V10/article/view/2134