Kaggle

I enjoy taking part at ML competitions on Kaggle. This page summarizes my achievements and provides links to blog posts, writeups and GitHub repos with my solutions. Check out my Kaggle profile to see more.

Highest user rank

🏆 Competitions: 293 out of 159,203 (Master)
📊 Notebooks: 330 out of 172,882 (Expert)
📁 Datasets: 75 out of 40,289 (Expert)
🗣 Discussion: 199 out of 196,413 (Expert)

Competitions

Computer vision

🥇 Cassava Leaf Disease Classification: top-1% [Summary] [GitHub]
🥇 SIIM-ISIC Melanoma Classification: top-1% [Summary] [Blog post]
🥈 BMS Molecular Translation: top-5% [Summary] [GitHub]
🥈 RANZCR Catheter and Line Position Challenge: top-5% [Summary] [GitHub]
🥉 Prostate Cancer Grade Assessment Challenge: top-6%
🥉 APTOS 2019 Blindness Detection: top-9% [GitHub] [Blog post]
🥉 RSNA STR Pulmonary Embolism Detection: top-13% [Blog post]

Tabular data

🥈 Google Analytics Customer Revenue Prediction: top-2% [GitHub]
🥈 IEEE-CIS Fraud Detection: top-3% [GitHub]
🥈 Home Credit Default Risk: top-4% [GitHub]
🥉 COVID-19 mRNA Vaccine Degradation Prediction: top-6%
🥉 Instant Gratification: top-6%
🥉 Mechanisms of Action Prediction: top-10%

Time series

🥈 PLAsTiCC Astronomical Classification: top-5% [GitHub]
🥉 Riiid! Answer Correctness Prediction: top-7%

Top-rated notebooks

🔥 LightGBM on Meta-Features: classifying pulmonary embolism with features extracted from X-rays
🔥 EfficientNet + Multi-Layer LSTM: translating molecule images to chemical formulas with deep learning
🔥 Stack Them All!: stacking ensemble pipeline for leaf disease classifiers