Sentiment & Reinforcement Learning
Pajamajoker / reinforcement-learning-for-kalshi-trading
What it does
DQN agent for hourly BTC threshold markets. Gymnasium-compatible RL environment; multi-day backtesting with policy comparison (random vs. baseline vs. DQN); Streamlit GUI.
What we learned from it
the RL framing for prediction markets is interesting but reward-shaping is hard. The author's policy comparison framework (always benchmark against a random baseline) is a useful discipline that many ML approaches skip.
Find the author
- Repository: github.com/Pajamajoker/reinforcement-learning-for-kalshi-trading
- GitHub profile: @Pajamajoker
- Website / blog: jpratham.info/
- Author bio: MS AI @ Katz, Yeshiva University | Ex Sr. SWE @ Intangles.ai | Ex SWE @ UBS
Don't want to maintain your own bot?
Bot for Kalshi is the no-code platform for automated prediction-market trading — visual builder, signals, encrypted credentials, kill switch. The infrastructure these open-source projects build, ready to use.