Home▸AI & Machine Learning▸3D Neural Collaborative Filtering versus Matrix Factorisation: When a Learned Interaction Beats the Dot Product

🎬 3D Neural Collaborative Filtering versus Matrix Factorisation: When a Learned Interaction Beats the Dot Product

Three Blender-built boards of bars, the true preference probabilities and the predictions of a dot-product matrix-factorisation model and a one-hidden-layer neural collaborative filter, trained live on three hidden taste worlds; compare held-out loss with the Bayes limit, gradient checks and parameter counts.

AI & Machine Learning3DAdvanced60 FPS📱 Mobile-adapted🟠 Blender⇄ 2D version
3D Neural Collaborative Filtering versus Matrix Factorisation: When a Learned Interaction Beats the Dot Product ↗ Open standalone
⚙ Under the hood

Three Blender-built boards of bars, the true preference probabilities and the predictions of a dot-product matrix-factorisation model and a one-hidden-layer neural collaborative filter, trained live on three hidden taste worlds; compare held-out loss with the Bayes limit, gradient checks and parameter counts.

Neural Collaborative FilteringMatrix FactorisationRecommender SystemsEmbeddingsImplicit FeedbackBackpropagationBayes ErrorMachine Learning

3D · Three.js / WebGL renderer · 60 FPS target · runs fully client-side, no install

What did you find?

Add reproduction steps (optional)