Machine-Learning-Based Semiparametric Time Series Conditional Variance: Estimation and Forecasting

This paper proposes a new combined semiparametric estimator of the conditional variance that takes product parametric and nonparametric based on machine learning. A popular kernel-based learning algorithm, known as kernel-regularized least squares estimator, is used to estimate component. We discuss how using real data use this make forecasts for variance. Simulations are conducted show dominance proposed in terms mean squared error. An empirical application S&P 500 daily returns analyzed, effectively future volatility.