Mini-project
IT, Software & SaaS
Game characters that learn: reinforcement learning agents
Train a game character with reinforcement learning, and compare it with one that follows hand-written rules.
The brief
Game characters that follow fixed rules are predictable. Build a small game or use a ready simulation, train an agent with reinforcement learning, and compare it with a rule-based character on score, fun and how easily a player beats it. Good looks like: a training curve, a fair comparison over many games, and a short playtest with five players.
What you will build
- Choose or build a small game environment
- Write a rule-based character as a baseline
- Train an agent with PPO or DQN
- Compare performance over 500 games
- Playtest with five players
- Write what learning added and cost
Free, open data to build it with
Sign up free to get the data and links
It takes a minute with your email. AiTeky Pass (₹999/month or ₹9,999/year) is free for your first 30 days.