A Comparative Study between Hierarchical Kernel Learning and Hybrid Models for Option Pricing

This paper delves into the realm of option pricing, presenting a comparative exploration of two distinct methodologies. Specifically, it focuses on the application of Hierarchical Kernel Learning (HKL) with two kernel functions: the Polynomial Kernel and the ANOVA Kernel. Furthermore, the paper introduces a hybrid models that combine Expectation-Maximization (EM) and K-Means clustering techniques with Support Vector Regression (SVR) . A comprehensive comparative analysis of the hybrid model sheds light on their capacity to improve option pricing accuracy across diverse market conditions. The insights derived from this research provide valuable contributions to understanding the performance of different kernel functions within the HKL framework. Moreover, it underscores the potential advantages of employing hybrid models in capturing nuanced market dynamics, thus enhancing the accuracy of option pricing. Ultimately, this study advances the field of option pricing, serving the needs of financial professionals and researchers seeking advanced tools for the valuation of derivatives.