Bachelor and Master Theses

Causal Machine Learning refers to the machine learning approach that models the data-generation process using a structural causal framework. Causal inference provides a mathematical foundation for modeling interventions and counterfactuals. However, integrating these principles into large-scale machine learning remains challenging due to issues of compositionality, abstraction, and scalability. On the other hand, Category theory is the branch of mathematics that deals with compositional structures. It provides a language for reasoning about complex systems by describing how parts combine to form wholes.

The aim of the thesis is to investigate category theory as a unifying mathematical foundation for causal machine learning. The task will be formalizing causal inference, i.e., structural causal models, interventions, and counterfactuals within the framework of category theory, which will allow compositional reasoning. The thesis will explore categorical abstractions to improve the design and understanding of causal machine learning.

The main research question is:

• How can category theory provide a compositional and interpretable foundation for causal reasoning in machine learning?

-       Deeper understanding on Graph Theory and Probability Theory

-       Good knowledge on Linear algebra, Basic Logic and algebraic thinking

-       Knowledge on functional programming using Python is an advantage