A system for monitoring statistical data distribution shifts in distributed settings
-
Updated
Dec 18, 2023 - Jupyter Notebook
A system for monitoring statistical data distribution shifts in distributed settings
Projet d'apprentissage des pratiques MLOps : pipeline complet de prédiction de prix immobiliers, du suivi d'expériences (MLflow) au serving (FastAPI), avec monitoring de dérive (Evidently) et CI/CD
Production-grade Employee Attrition Prediction system using FastAPI, Scikit-Learn, PyTorch Tabular ResNet/VAE, TreeSHAP explainability, KS drift detection, and What-If retention simulator.
This project builds a production-grade ML pipeline to classify Near-Earth Objects (NEOs) as hazardous or non-hazardous. It automates data ingestion, preprocessing, model training, monitoring, and drift detection using GitHub Actions, PostgreSQL, MLflow, DAGsHub, and Grafana.
pip package for data analysis stability evaluation against small data change.
To associate your repository with the data-drifts topic, visit your repo's landing page and select "manage topics."