II.
LibraryProcess overview
Reference · livelib-process:data-science-ml--model-monitoring-drift
model-monitoring-drift overview
Model Performance Monitoring and Drift Detection - Continuously monitor prediction accuracy, detect data drift and concept drift, track feature distributions, alert on performance degradation, and trigger automated retraining workflows with comprehensive quality gates.
Attributes
displayName
model-monitoring-drift
description
Model Performance Monitoring and Drift Detection - Continuously monitor prediction accuracy,
detect data drift and concept drift, track feature distributions, alert on performance degradation, and
trigger automated retraining workflows with comprehensive quality gates.
libraryPath
library/specializations/data-science-ml/model-monitoring-drift.js
specialization
data-science-ml
references
- - Evidently AI Model Monitoring: https://www.evidentlyai.com/ - Arize AI ML Observability: https://arize.com/ - WhyLabs AI Observability: https://whylabs.ai/ - MLOps Continuous Monitoring: https://cloud.google.com/architecture/mlops-continuous-delivery-and-automation-pipelines-in-machine-learning - Concept Drift Detection: https://arxiv.org/abs/1010.4784 - Data Drift in ML: https://towardsdatascience.com/understanding-dataset-shift-f2a5a262a766
example
const result = await orchestrate('specializations/data-science-ml/model-monitoring-drift', {
modelId: 'fraud-detection-v3',
monitoringWindowDays: 7,
driftThresholds: {
dataDrift: 0.15,
conceptDrift: 0.10,
featureDrift: 0.20
},
performanceThresholds: {
accuracyDrop: 0.05,
precisionDrop: 0.03,
latencyIncrease: 1.5,
errorRateIncrease: 0.02
},
alertChannels: ['slack', 'pagerduty', 'email']
});
usesAgents
- general-purpose
Outgoing edges
lib_applies_to_domain1
- domain:data-science·DomainData Science
lib_belongs_to_specialization1
- specialization:data-science-ml·Specialization
lib_implements_workflow1
- workflow:data-pipeline-deployment·WorkflowData Pipeline Deployment
lib_involves_role1
- role:data-scientist·RoleData Scientist
Incoming edges
None.