Anomaly Detection (Beta)
What is Anomaly Detection?
Anomaly detection is a technique used to identify unusual or abnormal patterns or behaviors within a dataset. In other words, it's like a detective work for data, where the algorithm is trained to spot data points that don't fit the normal or expected pattern or behavior.
Think of it like this - if you were a bank and you had a record of all your customers' financial transactions, you would want to identify any transactions that look suspicious, such as a large withdrawal from an account that has never had such a big withdrawal before. Anomaly detection algorithms can be trained to identify such unusual transactions automatically, helping the bank to detect potential fraud or errors.
In short, it is helpful when you want to identify unusual patterns or behaviors in your metric data that might indicate a problem, so that appropriate actions can be taken to address it.
PresightIQ Anomaly Detection
Based on historical values and fluctuation patterns of any metric, there is an expected range of future performance. When the metric actually performs outside of this range, Presight will spot this as an anomaly, and will highlight it to you via the means of a red dot.
Clicking on the dot, Presight will give you more information on the reason why we consider it anomalous. Clicking on Deep Dive will take you to a detailed panel.
After you've understood the reason for the detection, the next step is to let PresightIQ Auto-Insights guide you to exactly where the gist of the issue might lie.
Last updated
