The Finalist | Trappa
We made an LSTM machine learning model which classifies different financial fraud things into 8 consolidated classes with an accuracy of 70%. This was achieved by multiplying 100 initially given data points into 20K data points for model training by means of web search scraping.
Team name: Starstruck
Project name: Trappa
Team members: Demian Hetman, Artem Siryk, Ilya Kowalewski