Order Book Feature Engineering

Feature

This involves the transformation of raw order book data—bids, asks, and trade volumes at specific price levels—into quantifiable inputs for analytical models. Key derivations include measures of order book imbalance, liquidity pressure, and the effective spread across the top of the book. Creating robust features is essential for model generalization.