Google, IBM & Meta Certificates – 40% Off
One Coursera Plus subscription covers most Professional Certificates on Coursera.
Unlock All Certificates
This beginner path introduces reinforcement learning through the practical task of creating an intelligent agent that learns to navigate a grid-world environment. You will define states, actions, and rewards, then implement the environment that supports the learning process. You will build a Q-learning agent using the Bellman equation, Q-table updates, and an ε-greedy exploration strategy. You will connect the agent to its environment, train it over multiple episodes, and visualize its rewards and learned policy. The path also examines random goal positions, hazardous environments, and reward shaping. It concludes with a discussion of developing reinforcement learning techniques and their potential applications.
