Machine Learning in Finance: Balancing Randomness and Explainability
Dr. Bastian Bergmann, Executive Director, ETH FinsureTech Hub
Current Situation
The fascinating successes of Machine Learning (ML) in language processing, image recognition or multi-player games have triggered many fantasies to apply these technologies in other fields as well, including the area banking and finance. Typically, in the context of banking and finance, besides their impressive functionality basic questions on explainability and rationalization of the algorithms are in the focus.
In order to assess the potential of ML it is crucial for professionals to have both: A subject matter experience of ML techniques as well as a deep understanding of its scope and limitations. A particular emphasis will be given here to the fascinating role of randomness in ML procedures in stark contrast to classical technologies, where model parameters often have a meaning.
Objective
In this SFI Master Class we will go through some basic concepts of ML and its most common tools and programming techniques in view of latest research in banking and finance. In doing so, we will look at the concept of explainability and post-hoc rationalization, contrasting human decision-making with ML decision-making. Finally, we will put the concepts of testability and falsification, which proved to be a building block of scientific inquiry into perspective when using ML technologies.
Target Audience
The Master Class is aimed at everybody who wants to deepen her knowledge in machine learning and its potentials in modern financial industry. An interest in the conceptual underlying and in the philosophy of ML or AI is also welcome. Different backgrounds like economics, finance, or quantitative finance are welcome. No programming skills are required but one should not be afraid of discussing code.
SAQ Recertification
This Master Class is an acknowledged SAQ recertification measure for the CWMA profile and comprises four learning hours.