Global Coverage
Trained on multi‑region historical PM2.5 series with weather & satellite‑derived covariates for robust next‑day forecasts.
A compact summary of model runs and artifacts generated from our Kaggle notebook. Explore the notebook, review key metrics, and jump into the code.
Trained on multi‑region historical PM2.5 series with weather & satellite‑derived covariates for robust next‑day forecasts.
Gradient‑boosted trees + temporal features + geospatial embeddings; calibrated with cross‑validation and SHAP‑based insights.
Optimized preprocessing and batched prediction achieve sub‑second latency per 10k observations on commodity CPUs.
End‑to‑end notebook with fixed seeds, environment file, and exact artifact hashes for verifiable results.
Illustrative validation loss trend — replace with your chart or embed a PNG/SVG from the notebook.
Daily PM2.5 aggregates + meteorology (temp, wind, RH) + remote‑sensing proxies. Missingness handled via time‑aware imputation.
Lags/rolling stats (1–28d), holiday flags, temporal encodings, location embeddings; target leakage checks enforced.
XGBoost/CatBoost baseline → tuned via CV; error analysis with SHAP decision plots and partial dependence for sanity checks.
Temporal split validation; city‑wise breakdown; robustness to missing covariates; ablation on feature groups.