MD-ALL-SHAHRIA

MD-ALL-SHAHRIA/pulseair

Wearable air-quality risk forecasting with rigor-first evaluation — persistence floors, rolling-origin CV, GAN/SMOTE ablation, conformal prediction. Ships pulsebench, a standalone toolkit for evaluating imbalanced time-series models.

6 good first / help-wanted issues · Python · last activity Sep 22, 2026

0 stars 0 forks 0 watchers Python Other
air-quality benchmark class-imbalance conformal-prediction cross-validation ctgan edge-ai environmental-monitoring esp32 explainable-ai machine-learning model-evaluation onnx pytorch reproducible-research scikit-learn shap synthetic-data time-series uncertainty-quantification
View on GitHub Website
6 Open Issues Need Help Last updated: Sep 22, 2026

Open Issues Need Help

View All on GitHub
pulsebench: Holm-Bonferroni as a less conservative alternative in bonferroni_report about 1 month ago
enhancement good first issue pulsebench
MD-ALL-SHAHRIA/pulseair
0

Wearable air-quality risk forecasting with rigor-first evaluation — persistence floors, rolling-origin CV, GAN/SMOTE ablation, conformal prediction. Ships pulsebench, a standalone toolkit for evaluating imbalanced time-series models.

Python
#air-quality#benchmark#class-imbalance#conformal-prediction#cross-validation#ctgan#edge-ai#environmental-monitoring#esp32#explainable-ai#machine-learning#model-evaluation#onnx#pytorch#reproducible-research#scikit-learn#shap#synthetic-data#time-series#uncertainty-quantification
Transformer retraining hangs when the job is detached from the terminal about 1 month ago
bug help wanted
MD-ALL-SHAHRIA/pulseair
0

Wearable air-quality risk forecasting with rigor-first evaluation — persistence floors, rolling-origin CV, GAN/SMOTE ablation, conformal prediction. Ships pulsebench, a standalone toolkit for evaluating imbalanced time-series models.

Python
#air-quality#benchmark#class-imbalance#conformal-prediction#cross-validation#ctgan#edge-ai#environmental-monitoring#esp32#explainable-ai#machine-learning#model-evaluation#onnx#pytorch#reproducible-research#scikit-learn#shap#synthetic-data#time-series#uncertainty-quantification
Wanted: a multi-pollutant Dhaka ground-truth source (PM2.5 + PM10 + CO, hourly, 1+ year) about 1 month ago
help wanted methodology reproducibility
MD-ALL-SHAHRIA/pulseair
0

Wearable air-quality risk forecasting with rigor-first evaluation — persistence floors, rolling-origin CV, GAN/SMOTE ablation, conformal prediction. Ships pulsebench, a standalone toolkit for evaluating imbalanced time-series models.

Python
#air-quality#benchmark#class-imbalance#conformal-prediction#cross-validation#ctgan#edge-ai#environmental-monitoring#esp32#explainable-ai#machine-learning#model-evaluation#onnx#pytorch#reproducible-research#scikit-learn#shap#synthetic-data#time-series#uncertainty-quantification
Port the compressed RandomForest to C for the ESP32 (the ONNX benchmark is not a flashability claim) about 1 month ago
enhancement help wanted
MD-ALL-SHAHRIA/pulseair
0

Wearable air-quality risk forecasting with rigor-first evaluation — persistence floors, rolling-origin CV, GAN/SMOTE ablation, conformal prediction. Ships pulsebench, a standalone toolkit for evaluating imbalanced time-series models.

Python
#air-quality#benchmark#class-imbalance#conformal-prediction#cross-validation#ctgan#edge-ai#environmental-monitoring#esp32#explainable-ai#machine-learning#model-evaluation#onnx#pytorch#reproducible-research#scikit-learn#shap#synthetic-data#time-series#uncertainty-quantification
pulsebench: test coverage for non-hourly frequencies about 1 month ago
good first issue pulsebench reproducibility
MD-ALL-SHAHRIA/pulseair
0

Wearable air-quality risk forecasting with rigor-first evaluation — persistence floors, rolling-origin CV, GAN/SMOTE ablation, conformal prediction. Ships pulsebench, a standalone toolkit for evaluating imbalanced time-series models.

Python
#air-quality#benchmark#class-imbalance#conformal-prediction#cross-validation#ctgan#edge-ai#environmental-monitoring#esp32#explainable-ai#machine-learning#model-evaluation#onnx#pytorch#reproducible-research#scikit-learn#shap#synthetic-data#time-series#uncertainty-quantification
pulsebench: add a seasonal-naive baseline alongside persistence about 1 month ago
enhancement good first issue pulsebench
MD-ALL-SHAHRIA/pulseair
0

Wearable air-quality risk forecasting with rigor-first evaluation — persistence floors, rolling-origin CV, GAN/SMOTE ablation, conformal prediction. Ships pulsebench, a standalone toolkit for evaluating imbalanced time-series models.

Python
#air-quality#benchmark#class-imbalance#conformal-prediction#cross-validation#ctgan#edge-ai#environmental-monitoring#esp32#explainable-ai#machine-learning#model-evaluation#onnx#pytorch#reproducible-research#scikit-learn#shap#synthetic-data#time-series#uncertainty-quantification