The goal of this research is the development of a Curious Data Scientist Framework. This single framework is based on the curiosity loop architecture for learning data structures (feature and instance selection) in large and varied datasets. The repository currently include:
- Intrinsically motivated learning method for curious featrue selection (CFS)
- Intrinsically motivated learning method for curious instance selection (CIS)
- Scaling and generalization with deep learning architecture for curious featrue selection (DCFS)
- A single comprehensive feature selection-based clustering framework to create data substructures with features subset (CuFSC)