Kaggle Challenges
Jaguar Re-ID
Pairwise jaguar re-identification — scored as image similarity, not classification. The retrieval-first refactor moved a 0.421 public baseline to 0.871.
- result
- Leaderboard case study
- confidence
- High — public score and fixed validation both agree
- next
- Test higher-resolution EVA/Swin backbones only after preserving the retrieval validation split.
- Computer Vision
- Bioinformatics
- Audio
- Tabular
No catalog entries match the active filters.
Stanford RNA 3D Folding 2
Predict the 3D structure of RNA from sequence. A classical sequence-NN baseline beats anything that ignores the closest train neighbors; RibonanzaNet3D refines the candidates; a learned reranker picks the final five.
- result
- Active hybrid baseline
- next
- Tighten the learned reranker over the 128-candidate pool before more compute-heavy refinement.
BirdCLEF 2026
Acoustic species ID in Pantanal soundscapes — research-track Kaggle. Pipeline live with a class-prior baseline; deep audio modeling is next.
- result
- Submission-shape baseline
- next
- Confirm the official metric, then build grouped validation around soundscape windows.
AI Adoption · Fortune 500
EDA + clustering on a synthetic Fortune-500 AI-adoption panel. Random Forest ROI predictor, K-means personas, PCA projection, correlation atlas, use-case trajectories.
- result
- Synthetic EDA reference
- next
- Replace synthetic panel assumptions with a real longitudinal adoption source.
CSIRO Pasture Biomass Estimation
Multi-task ResNet50 + NDVI/Height fusion + constraint loss for 5 dry-matter targets on pasture imagery. 13% RMSE improvement over single-target baseline.
- result
- Archived case study
- next
- Run image-stratified cross-validation at 512px+ input resolution.
Student Performance EDA
EDA on student grades — Random Forest predictor of overall_score, behavioral K-means personas, sleep / study correlations, A→F grade distributions.
- result
- Synthetic EDA reference
- next
- Separate predictive features from policy levers before presenting causal takeaways.
Digit Recognizer
Legacy MNIST kept as the canonical scaffold — the smoke-test that proves challenge.json, the dashboard registry, the remote bootstrap, and the submit-to-Kaggle flow all still work.
- result
- Scaffold smoke test
- next
- Keep the run tiny and use it to verify auth, dashboard actions, and submit flow.
Titanic — Survival
Classic tabular starter — sex, age, fare, class. Stub catalog entry; full dossier pending.
- result
- Catalog stub
- next
- Add score progression + feature ablation writeup.
House Prices — Advanced Regression
Ames housing sale prices — heavy EDA + ensembling stub. Full writeup not started.
- result
- Catalog stub
- next
- Document ridge → LGBM → stack progression.
Spaceship Titanic
Synthetic passenger transport tabular — modern Titanic replacement. Stub only.
- result
- Catalog stub
- next
- Wire challenge.json + first LB submission.
Feedback Prize — English Language Learning
Essay scoring with disagreement-aware metrics — NLP tabular stub.
- result
- Catalog stub
- next
- Add trait-wise error breakdown chart.
Ventilator Pressure Prediction
Time-series breath curves → pressure — sequence model stub.
- result
- Catalog stub
- next
- Document inspiratory vs expiratory feature splits.
Smartphone Decimeter Challenge 2022
GNSS raw measurements → decimeter positioning. Stub catalog entry.
- result
- Catalog stub
- next
- Add map overlay of corrected trajectories.
Leaf Classification
Margin + shape + texture features on leaf scans — classic feature-engineering CV stub.
- result
- Catalog stub
- next
- Compare hand features vs small CNN baseline.
APTOS 2019 Blindness Detection
Retinal fundus images → diabetic retinopathy grade. Medical CV stub.
- result
- Catalog stub
- next
- Add confusion matrix + kappa progression table.