A reading list for people who want AI to trade their money

What it does

awesome-deep-trading is a curated reading list that catalogs academic papers, open-source repositories, datasets, and courses applying deep learning to algorithmic trading. The maintainer sorts research by technique—CNNs, LSTMs, GANs, reinforcement learning—and by financial domain such as high-frequency markets, cryptocurrency, and portfolio management. It is essentially a living literature review formatted as a GitHub awesome-list.

The interesting bit

The value is in the tedious taxonomic work: instead of dumping links, the maintainer separates limit-order-book papers from sentiment-analysis papers and distinguishes portfolio-management RL from scalping multi-agent systems. The repo also grants unusually permissive reuse rights, offering dual MIT/CC-BY licensing even for the curated list itself.

Key highlights

• Heavy focus on 2016–2020 academic literature, including dedicated sections for meta-analyses and systematic reviews.
• Organized by both architecture (CNN, LSTM, GAN) and trading context (high-frequency, portfolio, cryptocurrency, social/sentiment processing).
• Includes non-paper resources: guides, presentations, courses, and simulation datasets.
• Explicitly collects papers on market vulnerabilities and manipulation, not just profitable-sounding strategies.
• Dual-licensed under MIT or CC-BY.

Caveats

• This is pure curation, not code: you cannot clone it and run a backtest.
• The provided sources are truncated, so the full depth of the Repositories, Datasets, and Resources sections is unclear.
• Many entries link to publisher paywalls or third-party PDF hosts, so link rot is a predictable hazard.

Verdict

Useful as a jump-start for researchers, PhD students, or quant developers assembling a literature review on neural networks in finance. Look elsewhere if you need a turnkey trading framework or maintained backtesting library.