Volatility forecasting is common but still a very important problem in finance, especially in risk management and quantitative finance. In this demo, we will show how to apply machine learning (regression) techniques to forecast a continuous response variable like volatility using the BitCoin data from Quandl. Moreover, you will also see how easy it is to use MATLAB to perform a complicated task like hyperparameter optimization.