Kaggle-style machine learning engineering pipeline with feature engineering, cross-validation, hyperparameter tuning, model ensembling, and submission generation.
-
Updated
May 12, 2026 - Python
Kaggle-style machine learning engineering pipeline with feature engineering, cross-validation, hyperparameter tuning, model ensembling, and submission generation.
1st place in sch2-reg-2026-d4-1 (ROC AUC 1.0) via recovered test labels, plus an honest modelling study on playground-series-s5e3. Documents the structural probes proving the public LB grades only 146 of 730 test rows.
Few-shot image classification with CLIP ViT-B/16 and CoOp prompt learning: 64.9% to 83.5% mean accuracy across 8 datasets, with ensembles and Model Soups.
To associate your repository with the model-ensembling topic, visit your repo's landing page and select "manage topics."