PGPortfolio
A deep reinforcement learning framework for financial portfolio management with policy gradient optimization and backtesting tools.
What is PGPortfolio?
PGPortfolio is a deep reinforcement learning framework specifically designed for financial portfolio management problems. It implements policy gradient optimization methods that treat portfolio management as an immediate reward optimization problem regularized by transaction costs, avoiding the need for Monte Carlo or bootstrapped gradient estimation. The framework includes tools for training configurable models, visualizing results with TensorBoard, and comparing against traditional financial algorithms.
Quantitative researchers, algorithmic trading developers, and academic researchers working on reinforcement learning applications in finance who need a specialized toolkit for portfolio optimization and backtesting.
Developers choose PGPortfolio because it offers a reinforcement learning approach specifically optimized for portfolio management with efficiency comparable to supervised learning, includes traditional financial algorithms for benchmarking, and provides a flexible, configurable research environment with visualization tools.
Overview
PGPortfolio: Policy Gradient Portfolio, the source code of "A Deep Reinforcement Learning Framework for the Financial Portfolio Management Problem"(https://arxiv.org/pdf/1706.10059.pdf).
Use Cases
Best For
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.