FlowX流动矩阵-聚焦加密货币中的量化交易

When we use AI to achieve automated trading for us 😀
But we’re currently still losing money 😂
We use Deepseek as our large language model foundation to analyze BTC candlestick (K-line) data to determine when to buy. However, we’re still continuously losing money. I think it’s because our Prompt isn’t good enough. This project has been open-sourced; if you want to learn more, please visit Github
Liquidity Matrix - focusing on crypto quantitative trading

From Galois Lab's new open-source project
VolSignal builds indicator factors based on 1m K-line data, including return rate, moving average difference, abnormal trading volume, rolling volatility, RSI, MACD, Bollinger Bands, GK volatility, etc. Then it calculates the correlation between these factors and future returns/volatility, and uses them for subsequent machine learning model training.
#量化

Recently we shared a cryptocurrency volatility forecasting research project called VolSage. It is not a signal or “call” tool, and it’s not a bot that predicts whether prices will go up or down. Instead, it trains a model using Binance historical K-line (candlestick) data to try to determine the strength of a given coin’s volatility over a certain period in the future. Why do I think “volatility” is more worth studying than the “direction of price movement”? Because very often in trading, the first thing we need to judge isn’t whether it will go up or down, but rather: will the market become more turbulent next?

Disclaimer: This project is for learning purposes only and for sharing quantitative research. It does not constitute any investment advice. The cryptocurrency market carries extremely high risk. Please make your own judgments and assume the risk.

访问项目地址