Feature Engineering

The prediction models leverage a diverse range of feature categories to enhance their accuracy and robustness. These categories include:

• Technical Indicators: These are derived from the price and volume data of the asset, and they capture various aspects of the market's behavior, such as momentum, volatility, and trend. Examples of technical indicators include moving averages, relative strength index (RSI), and Bollinger Bands.

• Market Microstructure: These features provide insights into the underlying supply and demand dynamics of the market. They include order book data, trade data, and quote data. By analyzing these features, the models can identify potential buying or selling pressure, as well as other market inefficiencies.

• Temporal Features: These features capture the time-dependent aspects of the market, such as the time of day, day of the week, and seasonality. By incorporating temporal features, the models can account for the cyclical nature of the market and adjust their predictions accordingly.