- : Ms Word, Ms Word Format
- : 60 Pages
- : ₦10,000
- : 1-5 Chapters
THE IMPACT OF AUTOMATIC PARAMETER LEARNING METHOD FOR AGENT ACTIVATION SPREADING NETWORK BY EVOLUTIONARY COMPUTATION
Abstract:
The activation spreading network (ASN) is a computational framework used to model the behavior of agents in complex systems. The efficiency and effectiveness of an ASN depend on the proper configuration of its parameters, which can be a challenging task due to the high dimensionality and non-linearity of the parameter space. This study investigates the impact of an automatic parameter learning method based on evolutionary computation techniques for optimizing the performance of an ASN.
The proposed approach utilizes an evolutionary algorithm to evolve a population of parameter configurations, aiming to find the optimal set of parameters that maximizes the spreading efficiency and overall performance of the ASN. The fitness evaluation of each parameter configuration is based on a performance metric that captures the effectiveness of the agent activation spreading process.
To evaluate the effectiveness of the automatic parameter learning method, a set of experiments is conducted on various real-world and synthetic datasets. The results demonstrate that the evolutionary computation-based approach significantly improves the performance of the ASN compared to manual parameter tuning or random search methods. The optimized parameter configurations obtained through evolutionary computation consistently outperform other approaches, achieving higher spreading efficiency and better overall system performance.
Furthermore, the study investigates the impact of different evolutionary operators, such as selection strategies, crossover and mutation operators, on the performance of the automatic parameter learning method. The experimental results provide insights into the influence of these operators and their interactions, which can guide the design of more effective evolutionary algorithms for parameter optimization in ASN.
Overall, this research contributes to the field of agent-based modeling and computational social science by providing a systematic and automated approach for optimizing the parameters of an activation spreading network. The proposed automatic parameter learning method based on evolutionary computation offers a practical solution for improving the performance of complex systems modeled using ASN, enabling a more accurate representation of real-world phenomena and supporting decision-making processes in various domains.