ngiavng-art/MACHINE-LEARNING-MACD ? reverse-engineered prompt

Build me a Python project for automated crypto trading signal analysis based on MACD and machine learning. I want it to take market data, train and save models for different timeframes like 1m, 5m, and 15m, and use those models to help decide when to buy or sell. It should be able to work with existing saved models, keep logs, and store generated data in the right folders.

Please make it easy to run from a simple terminal or batch file, and include clear setup and usage steps in the project. If there are any missing details about the trading rules or data format, infer sensible defaults from the codebase and look up current docs online if needed. I also want the project to feel organized, with separate parts for data handling, model training, prediction, and tests.

Make sure it is practical for someone who wants to experiment with MACD based machine learning strategies, not just a demo.