Machine Learning and Prophecy of Behavior: A Breakthrough in Artificial Intelligence

Abstract

This paper addresses how the cognitive theory and machine learning methods are combined for predicting human behavior. Prediction of human behavior is a complex task. Researchers addressing this task proposes behavior models using statistical methods and rationality theory such as machine learning algorithms. This paper addresses how the potential cognitive theory and machine learning methods are combined in using by providing and predicting human behavior. We discussed two theory focal points theory and aspiration adaptation theory can be used in combination with machine learning algorithm to simplify the task and understanding the behavior. These two approaches are combined and compared in two domains. In simple domain, both machine learning models and psychological models give clear predictions and results. While in complex domains, the results states that psychological models’ results are not cleared and in learning models are also inaccurate. To overcome this problem, the hybrid model is introduced which helps in understanding of how people behave and at the same time how they wouldn’t behave reducing the space helping machine learning algorithms in searching the accurate behavior models. Hybrid model with cognitive models and machine learning have contributed in models which predicted behavior accurately as compared to machine learning models/cognitive models alone. The future researches must work in these directions using hybrid models for successful prediction of behavior.