LukeSantossz/weather-forecast

build(deps): pin dependency versions for reproducible builds about 2 hours ago

good first issue phase-0-hygiene tooling

0

End-to-end temperature forecasting pipeline, 0.19°C RMSE across 211 countries. Statistical models (ARIMA/SARIMA), gradient boosting (LightGBM), inverse-RMSE, weighted ensemble, and dual anomaly detection on 133K+ observations

Jupyter Notebook

#anomaly-detection#arima#data-science#forecasting#lightgbm#machine-learning#prophet#python#time-series#weather

docs: add LICENSE and CONTRIBUTING 17 days ago

documentation good first issue phase-0-hygiene

LukeSantossz/weather-forecast

0

End-to-end temperature forecasting pipeline, 0.19°C RMSE across 211 countries. Statistical models (ARIMA/SARIMA), gradient boosting (LightGBM), inverse-RMSE, weighted ensemble, and dual anomaly detection on 133K+ observations

Jupyter Notebook

#anomaly-detection#arima#data-science#forecasting#lightgbm#machine-learning#prophet#python#time-series#weather

fix(data_loader): validate required columns independently of date parsing 17 days ago

bug good first issue phase-0-hygiene

LukeSantossz/weather-forecast

0

End-to-end temperature forecasting pipeline, 0.19°C RMSE across 211 countries. Statistical models (ARIMA/SARIMA), gradient boosting (LightGBM), inverse-RMSE, weighted ensemble, and dual anomaly detection on 133K+ observations

Jupyter Notebook

#anomaly-detection#arima#data-science#forecasting#lightgbm#machine-learning#prophet#python#time-series#weather

fix(notebooks): strip hardcoded local paths and outputs from notebooks 17 days ago

bug good first issue phase-0-hygiene

LukeSantossz/weather-forecast

0

End-to-end temperature forecasting pipeline, 0.19°C RMSE across 211 countries. Statistical models (ARIMA/SARIMA), gradient boosting (LightGBM), inverse-RMSE, weighted ensemble, and dual anomaly detection on 133K+ observations

Jupyter Notebook

#anomaly-detection#arima#data-science#forecasting#lightgbm#machine-learning#prophet#python#time-series#weather

Open Issues Need Help