charliedream1/ai_quant_trade

★ 5,604Jupyter NotebookAudience · dataComplexity · 3/5Setup · moderate

Mindmap

mindmap
  root((ai_quant_trade))
    Strategy types
      Moving average rules
      Machine learning
      Reinforcement learning
      Graph neural networks
    Data tools
      Market data collection
      Factor mining
      News sentiment analysis
    LLM features
      Market report generation
      Price forecasting models
    Deployment
      Python and C++ support
      CPU and GPU targets
    Audience
      Beginners
      Experienced developers

mindmap root((ai_quant_trade)) Strategy types Moving average rules Machine learning Reinforcement learning Graph neural networks Data tools Market data collection Factor mining News sentiment analysis LLM features Market report generation Price forecasting models Deployment Python and C++ support CPU and GPU targets Audience Beginners Experienced developers

Click or tap to explore — scroll the page freely

Things people build with this

USE CASE 1

Learn and implement a moving-average crossover trading strategy in Python from a complete, explained code example.

USE CASE 2

Train a reinforcement learning agent to make stock buy and sell decisions and backtest it on historical price data.

USE CASE 3

Analyze financial news headlines with a large language model to generate automated daily market summary reports.

USE CASE 4

Connect example code to the JoinQuant online backtesting platform to run a live strategy test without building your own data pipeline.

Tech stack

PythonJupyter NotebookPyTorchpandasNumPy

Getting it running

Difficulty · moderate Time to first run · 1h+

In plain English

This is a Chinese-language learning and practice platform for AI-driven stock trading, aimed at both beginners with no programming background and experienced developers. The name translates roughly to "AI Quantitative Trading Operator," and the project covers the full journey from studying trading concepts to running strategies in a live market environment. The repository is organized as a large collection of example code and tutorials rather than a single deployable application. It covers many different approaches to automated trading decisions: traditional rule-based strategies like moving-average crossovers, machine learning models that look for statistical patterns in price data, deep learning networks, reinforcement learning agents that treat trading as a game of maximizing returns, graph neural networks that model relationships between stocks, and high-frequency trading techniques. One highlighted example claims a reinforcement learning strategy achieved 53% annualized returns in backtesting across multiple stocks. Beyond trading strategy code, the project includes tools for collecting and processing stock market data (including integration with the Wind financial data platform), factor mining (automatically discovering large numbers of statistical signals that might predict price movement), and financial text analysis that reads news or social content to gauge market sentiment. A newer section covers using large language models for tasks like generating financial market reports and training specialized models for stock price forecasting. For those who want to trade without building strategies from scratch, there are auxiliary tools: a live-market monitoring display built in Excel, a stock recommendation module, and example code for connecting to an online backtesting platform called JoinQuant. Deployment guides cover running strategies in Python or C++ on both CPU and GPU hardware. The core repository is free to use, but the maintainer also runs a paid community where additional tutorials, private code updates, and video courses are sold. All examples include explanations of the underlying concepts alongside the code, with the goal of letting someone with no finance or programming background work through the material step by step.

Copy-paste prompts

Prompt 1

Using the ai_quant_trade reinforcement learning example, help me build a Python trading agent that learns to buy and sell a stock to maximize returns on historical price data.

Prompt 2

Show me how to implement the moving-average crossover strategy from ai_quant_trade and backtest it on a CSV file of daily closing prices.

Prompt 3

Help me use a large language model to generate a daily market summary report from financial news headlines, following the LLM reporting section of ai_quant_trade.

Prompt 4

I want to add a sentiment analysis step to a trading strategy that reads recent news and scores it positive or negative before placing a trade. Help me set that up using ai_quant_trade as a reference.

Open on GitHub → Explain another repo

← charliedream1 on gitmyhub — every repo by this author, as a profile.

Verify against the repo before relying on details.