Master theses

If you plan to write your master thesis at the Department of Business Mathematics and Data Science, please contact Professor Dimpfl by mail.

Topics will be assigned upon mutual agreement. To find a suitable topic, please follow the content of the lectures and the research of the department. When you have an idea, please formulate a brief topic proposal (no more than two pages) including at least two articles relevant for the topic. As an inspiration for a possible topic, the list of past topics provided below might be helpful.

Topics of completed master theses

Stock Price Prediction using LSTM for a High-Frequency Trading Strategy
Basic literature: Arévalo et al. (2016) High-Frequency Trading Strategy based on Deep Neural Networks
The Development of Market Efficiency in a Cryptocurrency Market
Basic literature: Rösch et al. (2016) The Dynamics of Market Efficiency
From Mean to Machine Learning to Consumption-Based Asset Pricing - A Model Performance Analysis on the Tennis Sports Betting Market
Surprise Surprise - Using Deep Learning Methods to Model Emotional Cues During Soccer Matches with Incomplete Data
Modeling Stock Market Intraday Seasonality: Machine Learning vs. Traditional Approaches
Basic literature: Bauwens und Giot (2000) The Logarithmic ACD Model: An Application to the Bid-Ask Quote Process of Three NYSE Stocks

Formalia

The processing time is four months.
The master's thesis can be written in German or English..
For the production we recommend the scientific text typesetting program Latex. All the software required for this is freely available.
- zip file with all raw files and template
- compiled document (with notes on the software and guidelines for formal requirements)
- A detailed instruction video on Youtube
The scope of the work is 11000-12000 words, i.e. approx. 25-30 pages of text. There are also graphics, a bibliography, etc. The total volume should not exceed 45-50 pages.