Applied Artificial Intelligence Tasklab
Anomaly Detection
Is a technique used in statistics to identify data points that deviate from the expected behavior of a system. Using AI techologies this is can be used systematically and extensively analyzing patterns and identifying outliers that may indicate unusual ocurrences.
How the task is defined
- Input
- Data streams or datasets
- Output
- Identification of outliers or anomalous instances
- Goal
- To detect unusual patterns that do not conform to expected behavior
- Learning strategy
- Statistical, clustering, or neural network models
- Evaluation metric
- Precision, recall, F1 score, and detection time