Published On Jan 28, 2021
Suck at games?
Keep getting beaten?
Want to absolutely smash everyone at Space Invaders?
Oh, have I got a solution for you! In this video you'll see the journey to building an AI model that's able to crush it at Space Invaders. We'll use, Reinforcement Learning, a similar technique to what was used by DeepMind in their documentary on AlphaGo.
The model shown in the video was built using an ACER agent with OpenAI's stable_baselines package and Python. After around 40m steps the model performance tailed off a bit so the final model shown was one that seemed to consistently perform reasonably well.
Oh, and don't forget to connect with me!
LinkedIn: / nicholasrenotte
Facebook: / nickrenotte
GitHub: https://github.com/nicknochnack
Happy coding!
Nick
P.s. Let me know how you go and drop a comment if you need a hand!