portfolio-management

Inspiration

we wanted to know can reinforcement learning learn to manage a portfolio, or will it simply exploit patterns that work in certain regimes

What it does

tests RL against versions of Hidden Markov Model's and a simple BUY and HOLD strategy

How we built it

Challenges we ran into

"included in the video and report.md"

Accomplishments that we're proud of

we were able to pinpoint the problem with RL, iterated through many versions of HMM's till we found what works

What we learned

applying a Modern day machine-learning paradigm to trading "at least" is more challenging than what we thought

What's next for portfolio-management

definitely tweaking the reinforcement learning approach or testing other machine-learning paradigms

Built With

• gymnasium
• kaggle
• python
• pytoch
• yfinance