Projects

Train a Smartcab to Drive with Reinforcement Learning

• Designed a Q-learning algorithm in Python that trained a simulated car to follow US traffic rules. • Implemented learning rule to take effect within 28 trials with reliability of over 80%.