Volatility in the stock market captures price fluctuations and is therefore an important quantitative indicator in financial
markets. In this project, we aim to use order book and trade data to predict the Realized Volatility (RV) over the immediate
next 10 mins. The universal loss function used throughout the project is the Root Mean Squared Percentage Error
(RMSPE). This project is based on the Optiver Realized Volatility Prediction Kaggle competition (2021).