Please enable JavaScript.
Coggle requires JavaScript to display documents.
ML Ops 2022 - Coggle Diagram
ML Ops 2022
Manage
Model Management (Alan)
Model Documentation (Model Cards) (Alan)
Model Security (2023 - Adversarial attack prevention)
Sunset / Support Strategy (George)
Run / Enhancement / Project (George)
Feature store (Brendon)
Data Pipeline and Coding Standards (Sai, Brendon)
Pipeline Lineage (Brendon)
Monitor
Data Drift Detection (Dian, Brendon)
Model Quality
Prediction Performance (Dian)
Covariate Shift = shift in target variable
Prior Probability Shift = shift in the relationship between independent and target variables
Concept Drift = Shift in the independent variables
Bias Detection (Dian)
Usage
Model Retraining
Data Science Infrastructure Monitoring (George) (2023)
Deploy
Batch
Automated
On-Demand
Variation in data readiness
Human review required prior to publish
Transactional (Sai)
Persistent Compute
Serverless Compute
Data Science Containers (Sai)
Assess
Experiment Logging (Darrell)
Auto ML Expansion (Darrell)