Faking statistics or doing bogus research on data has always been a classic and interesting topic. In the big data age, we observe otherwise rare phenomena such as the Simpson’s paradox more often. There are also limits to our methods, both theoretical - think “black swan” - and human - think biases. I want to touch several topics to increase your consciousness and sharpen your critical thinking as an ethical data scientist. As everyone in Machine Learning has created a faulty experimental design at least once, this presentation is also of a high practical value. I will show-case you concrete examples of where the model evaluation has been screwed up for the disadvantage of human beings.

A seminar paper for the “Current Issues in Macroeconomics” seminar at the FernUniversität in Hagen. The paper explains various methods for measuring inflation expectations (IE), including survey-based (households, firms, professional forecasters) and market-based approaches. It also analyzes their accuracy, uncertainty, and the concept of “anchoring” of expectations in the context of central bank credibility.

A seminar paper written for the course “Environmental Economics” at the FernUniversität in Hagen. It analyzes, based on a microeconomic model by Kennard (2020), why some firms support costly climate change legislation while others oppose it. The core idea is that firms with a “green” capital advantage may lobby for stricter regulations to gain a competitive edge over rivals with higher adjustment costs.

Note: Paper temporarily unavailable while sensitive data is being removed.

Figures in the Euclidean plane $\mathbb{R}^2$ can be generalized to have orientiation. In the case of a polygon this means that the edges are directed. This allows to introduce notions usually known from differential topology such as the winding number or the degree of a figure. Interesting theorems relating these quantities are discovered and generalized operations (convolution, intersection of such directed polygons) introduced. This framework leads to new insights regarding problems of computational geometry and to new algorithms such as the convolution method to calculate Minkowski sums of two polygons.

The (co-)homology groups of groups are introduced in terms of universal functors coined by Grothendieck. In the case of finite groups $G$ the sum of all elements $\prod_{g \in G} g$ exists and induces a map between the $0$-dimensional group homology and group cohomology which leads to a unified cohomology theory (the Tate cohomology) that has nice properties. Finally results from this theory are applied to the theory of local fields (following work of Serre). _With Klaus Altmann, first of two mentors of the field medalist Peter Scholze._

Introduction to the field of cryptography and its relation to modern complexity theory. (Warning: end of proof of theorem 3 is obviously wrong and can be fixed easily).

A survey regarding security in sensor networks. Introduces basic notions and common attack vectors. Talks more in depth about stream ciphers that are useful for this kind of setting.

Report on how to do common attacks on wireless networks with the BackTrack Linux distribution. Jointly with Philipp Breinlinger.

My contribution to a seminar about (topological) geometric group theory where I presented quasi-isometries.

A short paper that introduces the group structure of points on an elliptical curve.

An introduction to homological algebra including exact sequences, the hom functor, and the snake lemma. Jointly with Oliver Schulze.

In a graph theory seminar we presented about extremal graph theory as a group. My part was about the regularity lemma of Szemerédi. Jointly with Dr. Laura Gellert (and some others I forgot).

My two-page summary of the lecture “Höhere Algorithmik” (Higher Algorithmics) at FU Berlin that could be taken into the examination.