Success lies outside of your comfort zone.

A quote (as cheesy as it sounds) that I always try to implement in my life. Be it through constantly learning something new or with my decision to pursue research.

Biography

I did my Bachelor of Science in E-Commerce (focus on Computer Science) in Germany in a dual studies programme at the FH Wedel – University of Applied Sciences. That meant I worked and studied in alternation giving me the opportunity to use what I have learned in university directly into a practical field in a company.

For my Master of Science in Business Analytics I went to the University of Edinburgh in beautiful Scotland. My Master’s dissertation (it is indeed called ‘dissertation’ in the UK) was called Cluster Analysis for Customer Segmentation with Open Banking Data and dealt with finding customer segments using transaction data from Open Banking Data in the UK.

Research Interests

My research interests lie in Machine Learning and Deep Learning Applications for Computer Vision.

Publications

You can find my very first publication here:

Catja Bartels. 2022. Cluster Analysis for Customer Segmentation with Open Banking Data. In 2022 3rd Asia Service Sciences and Software Engineering Conference (ASSE’ 22). Association for Computing Machinery, New York, NY, USA, 87–94. https://doi.org/10.1145/3523181.3523194

Portfolio

Here are my latest research projects:

Cluster Analysis for Customer Segmentation with Open Banking Data

I classified three customer groups representing the most, least, and medium valuable customers by using the RFM-Model and the clustering algorithms K-Means and DBSCAN on bank transaction data.
Paper: https://doi.org/10.1145/3523181.3523194
Code: https://github.com/CatjaB/Cluster-Analysis-for-Customer-Segmentation-With-Open-Banking-Data

  • Data: I used data sets from Open Banking, which include anonymized transactions (mainly focusing on different expense types) from various bank customers in the UK for three months in 2017.
  • Methods: I used the Recency, Frequency, and Monetary Value (RFM) Model, and then the clustering algorithms K-Means and Density-Based Spatial Clustering of Applications with Noise (DBSCAN) with K-Means performing better than DBSCAN.
  • Result: I classified three customer groups representing the most, least, and medium valuable customers, based on their average values for different spending types.
PCA-Graph for explained variance for January
Scatterplot of K-Means with PCA for January

Contact

Feel free to contact me through LinkedIn or per mail: first name DOT last name AT outlook DOT com