This is the repository of the submitted solution for Ristek's Datathon 2024 by Ristek Universitas Indonesia. Submitted by Kuliah di Depok team and won first place out of 316 other competing teams.
The members of the team:
The main idea of this solution is to utilize hyperdimensional computing to elevate faster retrieval methods that can acquire context.
Initialize product vector
python init.py n_dimension
Validate model performance with Leave-One-Group-Out Cross Validation
python test.py n_dimension