How asebaie/triple-i works
Triple-I is a collection of independent data-science and machine-learning projects, not one application with a shared runtime workflow. The README documents each project’s goal and, for some, its data or prediction task; no source excerpts are available to confirm internal implementation or calls. The graph therefore shows documented project capabilities and inputs only, without inferring internal stages.
Components 9
Banking AI
- Income and recommendations
Recommendations
- Autoencoder recommender
Molecular ML
- Cocrystal classifier
Spectral ML
- Hyperspectral classification
Search Analytics
- Hybrid structured search
Product Matching
- Product-pair classifier
Other components
- Netflix Prize data
- Molecular SMILES pairs
- Ozon product-card pairs
Connections 3
- Netflix Prize data → Autoencoder recommender: trains on
README.md - Molecular SMILES pairs → Cocrystal classifier: classifies pairs
README.md - Ozon product-card pairs → Product-pair classifier: classifies pairs
README.md