[email protected] | |
3275638434 | |
Paper Publishing WeChat |
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License
MarioDell′Era
Full-Text PDF XML 475 Views
DOI:10.17265/1537-1514/2022.03.003
London,United Kingdom
In quantitative finance, numerical methods are commonly used for the valuation of financial quantities. Derivative price models are often multi-dimensional and moreover, closed-form solutions are not available. Over the time different numerical methods have been developed and introduced in literature, to solve partial differential equations (PDEs) or integral equations. To this list of numerical approaches, in the recent literature, appears more often the use of Neural Networks. The mathematical foundation that allows the Neural Networks to approximate financial quantities and not only, is the “Universal Approximation Theorem”. In this paper we are going to introduce the Stochastic Hopfield Neural Network (or Boltzmann machine) to estimate the Local Volatility σ(St, t), in order to speed up the calibration process, which often is time consuming and becomes a key problem to overcome.
Quantitative Finance, Option Pricing, Local Volatility, Numerical Methods, Neural Networks