Option Analysis With Advanced Artificial Intelligence Techniques

In this Thesis, we will conduct a comprehensive analysis of the topic of option pricing. We will start with a detailed analysis of this derivative (an option), as well as of classical methods for calculating the theoretical value of options (Binomial model, Black-Scholes model). At this stage, we will explain why analytical solutions (explicitly represented in the form of formulas) or differential equations solved by numerical methods have several limitations. Then we will introduce our own assumptions and explain why the use of Artificial Intelligence methods has several advantages and fewer limitations. Next, we will choose the optimal model and interpret the results obtained. The practical benefits and novelty of the research are explained, among other things, by the openness of data and code, as well as a data-oriented approach. Our focus is not only to find the most sophisticated model and architecture, but also to understand which factors explain the dynamics of option prices. Therefore, in addition to modeling, much attention will be paid to data analysis, feature engineering, and statistical hypothesis testing. Finally, we will touch on the limitations of our research, and explain possible further steps to improve it.