Anomaly Detection
summary
Workshop previews a concrete anomaly workflow: use Python, pandas and the TDDa library to write concrete constraints (uniqueness, min/max, allowed values), run tdda detect to spit out offending rows and get detected 33 bad records from a 15M-row price set, and run tdda discover to discover constraints from good data that can be tuned. Slides and exercises show starter code, command-line examples and a practical pattern/seasonality detector that actually flags and quantifies breaches.
AI-written summary · about
chapters (11)
- 0:00Anomaly detection concepts
- 3:00What makes data anomalous
- 8:20Simple anomaly detection methods
- 13:15Visualizing anomalies
- 20:21Balancing false positives
- 28:38One-dimensional detection with constraints
- 34:08Hands-on anomaly detection exercise
- 44:14TDD for automated detection
- 53:57Generating constraints with TDD
- 1:04:26Handling seasonality and trends
- 1:11:54Time-series anomaly detection
related talks
45mAnomaly Detection in Time Series
22mAnomaly Detection: The Bad Boy of Data Science
44mADTK: An open-source Python toolkit for anomaly detection in...
18mCross-Industry Anomaly Detection Solutions with R and Shiny
30mBetter data quality with ML & visualization
19mThe Rise of AI in Anomaly Detection- 18mCross-Industry Anomaly Detection Solutions with R and Shiny
29mREAL-TIME ANOMALY DETECTION MADE EASY