The project addresses a critical challenge in the field of machine learning by exploring innovative approaches to enhance autonomous decision-making in situations where information is incomplete or uncertain. Traditional reinforcement learning (RL) algorithms often struggle when faced with imperfect information, leading to suboptimal decision-making and limited real-world applicability. This project aims to advance the capabilities of RL systems by introducing self-reinforced learning mechanisms specifically designed to thrive in environments with incomplete or uncertain data. The core objective of the project is to develop a novel self-reinforcement framework that enables agents to adapt and improve their decision-making processes over time, even in the absence of perfect information. The research leverages insights from cognitive science, game theory, and information theory to design agents capable of self-assessing and refining their strategies based on the evolving nature of the environment.