High Interaction Multi-Agent System Model for Automatic Prediction

Abstract

In a cooperative multi-agent system, also known as a MAS, the behaviors of several agents are coordinated with one another so that they may work together to achieve a common goal, such as the completion of a task or the maximization of utility. As a consequence of this, there has been a surge in enthusiasm for applying techniques of machine learning to the task of automating the search and enhancement that is necessary when attempting to code answers to MAS problems. This is because these techniques can improve the accuracy of the search results. For this reason, we provide an interactive multi-agent model exploiting three different machine learning models that can predict the cost of a smartphone. A smartphone dataset was collected from Kaggle, and it was used in an investigation on the efficacy of the tactics that were recommended. The results of the experiments yield a prediction accuracy of 95% and a decision accuracy of 100%, demonstrating that a multi-agent system that learns may produce more accurate predictions than approaches that are currently considered state-of-the-art.