CS-370

This project implemented a reinforcement learning solution (specifically, deep Q-learning) to solve a treasure maze game. The intelligent agent learns to navigate the maze, find the treasure, and avoid obstacles.

My Work

• Core Implementation:
  • Developed the Q-learning algorithm within the qtrain function.
  • Managed the agent's state representation, decision-making (exploration vs. exploitation), and Q-value updates.
  • Integrated and trained the neural network model.
• Hyperparameter Optimization: Experimented with parameters such as exploration rate, experience replay memory size, and training data size to improve the agent's performance.
• Success Evaluation: Defined win conditions and completion criteria to track the agent's progress.

Understanding Computer Science

• Computer Scientists' Role: Computer scientists design algorithms, build systems, and drive technological innovation to solve real-world problems across various sectors.
• Importance: Computer science is the backbone of the digital world, influencing how we interact, work, and learn.
• The Problem-Solving Process
  • Decomposition: Break down complex challenges into smaller ones.
  • Abstraction: Focus on the big picture while managing details.
  • Algorithmic Thinking: Design precise instructions for problem-solving.
  • Iterative Development: Continuously test, evaluate, and refine solutions.
• Ethics in AI
  • User Well-being: Prioritize safety, fairness, and the avoidance of bias.
  • Transparency and Accountability: Be clear about how AI is used and align it with organizational values to foster trust.