In my thesis for Business Data Science I deploy advanced natural language processing (NLP) techniques to extract features from text data and consequently use these features in predictive modeling. My supervisor is prof. Marc van de Wardt (Business Data Science). On this page I briefly log my trajectory. In my thesis for Data Science & Marketing Analytics I focus on enhancing the detection of outliers and I am supervised by prof. Philip Hans Franses (Econometric Institute) and prof. Eran Raviv (Business Data Science).

Currently, I am the Teaching Assistant for Deep Learning, a graduate-level course of the Research Master’s program in Business Data Science at the Tinbergen Institute.

Previously, I was a Teaching Assistant for the following courses:

• Deep Learning (Summer School), graduate-level course at the Tinbergen Institute, Teaching Assistant to prof. Raviv.
• Principles of Economics & Business, undergraduate-level course at the University of Amsterdam, Teaching Assistant to prof. Schinkel.

Currently, I am working on three interesting projects. The first one is a continuation of my thesis at Business Data Science, in which I created text-based measurements of personality scores from web-scraped parliamentary speeches. The results indeed show that machine encoded features outperform conventional human encoded features (e.g. LIWC, MRC) in predicting personality scores, and that text-based measures of personality traits can serve as an appropriate proxy for survey-based measures of personality traits. The second project aims to find a novel method in alleviating stress. The data is collected by means of daily measurements of perceptions of stress, and these are consequently combined with physical indicators collected through wearable devices. The third project aims to efficiently (1) summarize and (2) synthesize large corpora of text. These automated literature reviews will initially be developed in the context of scientific literature in a specific subfield of Management Science: Careers.