Time series anomaly detection
Time series anomaly detection, including: gathering data associated with a particular event type and a particular user; generating, based on the data, a time series analysis; detecting an anomaly based on the time series analysis; and generating information describing the anomaly.
1. A method of time series anomaly detection, the method comprising:
gathering data associated with a particular event type and a user;
generating, based on the data, a time series analysis;
detecting an anomaly based on the time series analysis;
generating information describing the anomaly, wherein the information comprises an observation record that is one of a plurality of observation records; and
presenting a user interface describing a degree of satisfaction, based on the plurality of observation records, of a plurality of conditions comprising a top-level condition and one or more dependent conditions required to satisfy the top-level condition.
2. The method of claim 1 , wherein the information describing the anomaly comprises an alert.
3. The method of claim 2 , further comprising presenting the alert in a polygraph in which nodes are applicable logical entities, and edges represent behavioral relationships between the logical entities in the polygraph.
4. The method of claim 1 , wherein the time series is further based on data associated with the particular event type and one or more other users.
5. The method of claim 1 , further comprising adding an amount of non-zero noise to the data.
6. The method of claim 1 , wherein the user interface comprises a polygraph indicating one or more entities associated with one or more observation records of the plurality of observation records applicable to the plurality of conditions.
7. The method of claim 1 , wherein the user interface comprises a list describing one or more observation records of the plurality of observation records applicable to the plurality of conditions.
8. The method of claim 7 , wherein the user interface comprises a plurality of selectable elements each corresponding to the one or more dependent conditions, and wherein selecting a selectable element of the one or more dependent conditions filters the list to a subset of the plurality of observation records applicable to the dependent condition corresponding to the selected element.
9. A computer program product for time series anomaly detection, the computer program product disposed on a non-transitory computer readable medium, the computer program product including computer program instructions configurable to carry out the operations of:
gathering data associated with a particular event type and a particular user; generating, based on the data, a time series analysis;
detecting an anomaly based on the time series analysis; generating information describing the anomaly, wherein the information comprises an observation record that is one of a plurality of observation records; and
presenting a user interface describing a degree of satisfaction, based on the plurality of observation records, of a plurality of conditions comprising a top-level condition and one or more dependent conditions required to satisfy the top-level condition.
10. The computer program product of claim 9 , wherein the information describing the anomaly comprises an alert.
11. The computer program product of claim 10 , wherein the operations further comprise presenting the alert in a polygraph in which nodes are applicable logical entities, and edges represent behavioral relationships between the logical entities in the polygraph.
12. The computer program product of claim 9 , wherein the time series is further based on data associated with the particular event type and one or more other users.
13. The computer program product of claim 9 , wherein the operations further comprise adding an amount of non-zero noise to the data.
14. The computer program product of claim 9 , wherein the observation record includes various attributes of the anomaly including a time at which the anomaly occurred, one or more metrics associated with the anomaly including a predicted value or an observed value used to detect the anomaly, and identifiers for one or more compute resources or machines associated with the anomaly.
15. The computer program product of claim 9 , wherein the user interface comprises a polygraph indicating one or more entities associated with one or more observation records of the plurality of observation records applicable to the plurality of conditions.
16. The computer program product of claim 9 , wherein the user interface comprises a list describing one or more observation records of the plurality of observation records applicable to the plurality of conditions.
17. The computer program product of claim 16 , wherein the user interface comprises a plurality of selectable elements each corresponding to the one or more dependent conditions, and wherein selecting a selectable element of the one or more dependent conditions filters the list to a subset of the plurality of observation records applicable to the dependent condition corresponding to the selected element.
18. A system comprising:
one or more processing resources; and
a non-transitory computer-readable medium, coupled to the one or more processing resources, having stored therein instructions that when executed by the one or more processing resources cause the one or more processing resources to:
gather data associated with a particular event type and a user;
generate, based on the data, a time series analysis;
detect an anomaly based on the time series analysis;
generate information describing the anomaly, wherein the information comprises an observation record that is one of a plurality of observation records; and
present a user interface describing a degree of satisfaction, based on the plurality of observation records, of a plurality of conditions comprising a top-level condition and one or more dependent conditions required to satisfy the top-level condition.
19. The system of claim 18 , wherein the information describing the anomaly comprises an alert.
20. The system of claim 18 , wherein the instructions that when executed by the one or more processing resources further cause the one or more processing resources to:
present the alert in a polygraph in which nodes are applicable logical entities, and edges represent behavioral relationships between the logical entities in the polygraph.