Sentiment & Reinforcement Learning

Pajamajoker / reinforcement-learning-for-kalshi-trading

by Prathamesh Joshi (Pajamajoker)

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.

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.

Start Free Trial →