stefan-jansen/machine-learning-for-trading ? reverse-engineered prompt

Build me a comprehensive Python project that explores using machine learning for algorithmic trading. I want it to be a collection of educational Jupyter notebooks, maybe organized like chapters in a book, so I can learn from and experiment with the code.

It should cover the whole workflow, from sourcing market data and alternative data (like financial news from SEC filings) to engineering features and building predictive models. I want to see a wide range of techniques, from simpler models to advanced deep learning. Please include examples that show how to use AI to analyze text for sentiment, and how to apply different kinds of neural networks for forecasting.

A really important part is designing and backtesting the actual trading strategies based on the model predictions. I need to see how to evaluate their performance. Also, let's explore some more advanced topics, like generating synthetic stock data for training and even building a trading agent with deep reinforcement learning. You can use standard libraries like pandas and TensorFlow to build this out.