A learning algorithm for playing 2048
- 04-10:
- Extract score from game
- Extract states
- Extract terminal states (win/lose)
- Create interface with game
- Input structure
- Restart game
- 04-21:
- Create the Deep Q-Agent
- Experiment with different architectures
- Evaluate score โ Select most promising Architecture
- 05-20:
- Write Report and Presentation
- 05-24: Deadline