Menü
The simulated data comes from transactions between thousands of customers (a total of 11,000) in a mobile banking network. Most users behave according to a statistical model that describes normal behavior. However, some of them commit fraud.
The goal is to identify the fraudulent accounts based on their transaction behavior. For the Student Track an initial set of features is provided, but discovering new, more informative features is additionally rewarded. The fraudster populations in the student and in the Professional Track differ, so both tracks offer different challenges.
| Rank | Score | Team Name | Member Name(s) |
|---|---|---|---|
| 🏆1 | 99.83% | Import teamName | Ayden Janssen, Robyn Sophie Kruse |
| 🥈2 | 97.05% | Data Alchemy | Tobias Fricke |
| 🥉3 | 88.95% | :^) | Arne Winter |
| 4 | 85.64% | DataForcer | Abhirup Sinha, Pritilata Saha |
| 5 | 82.15% | BitBubenBande | Jann Eike Ackermann, Joshua, Aning, Max Lehmann |
| Rank | Score | Team Name | Member Name(s) |
|---|---|---|---|
| 🏆1 | 99.92% | Spider Bobs | Ali Salehzadeh-Yazdi, Eda Cakir, Johannes Falk |
| 🥈2 | 99.76% | Import teamName | Ayden Janssen, Robyn Sophie Kruse |
| 🥉3 | 99.61% | OldSchool | Yale Hartmann |
| 4 | 92.14% | Kornstante | Fabian Wetjen, Wilhelm Jochim, Marcel Plutat |
| 5 | 76.68% | IntroToAML | Eike Voß, Tom Splittgerbe |
