Anomaly/Outlier Detection

Anomaly/Outlier Detection

Application Description:

Consists of detecting anomalous behavior of an entity while conducting a transaction or performing some activity

  • “The problem of finding patterns in data that do not conform to expected behavior”

What’s different

  • Deliver a robust data science framework to detect anomalous events
  • Ingest and process large amount of data to score anomalous behavior to enable appropriate actions
  • Detect varied patterns of anomalies using training
  • Carry out appropriate processing based on time series data or textual data
  • Adopt latest technologies based on Deep learning if required by the problem
  • Perform detailed analysis and validation to ensure the anomalies identified can be trusted
  • Involve business in the iterative process to increase the relevance of the findings

Related Projects