| CPC G06F 21/554 (2013.01) [G06F 2221/034 (2013.01)] | 20 Claims |
|
1. A system for determining monitoring parameters within datasets and detecting anomalies within the datasets transformed into timeseries data, the system comprising:
one or more processors; and
a non-transitory computer-readable storage medium storing instructions, which, when executed by the one or more processors, cause the one or more processors to perform operations comprising:
receiving a base dataset, wherein the base dataset comprises a plurality of features and a plurality of entries;
determining a target feature of the base dataset;
inputting the base dataset and the target feature into a machine learning model to obtain one or more auxiliary features, wherein the machine learning model selects the one or more auxiliary features based on how well the one or more auxiliary features enable segmentation of the base dataset in relation to the target feature;
providing, to a monitoring system, the target feature and the one or more auxiliary features;
receiving, from the monitoring system, a monitoring dataset for the target feature, wherein the monitoring dataset comprises a first plurality of values for the one or more auxiliary features and a second plurality of values for the target feature;
segmenting and aggregating the target feature into a plurality of timeseries datasets based on an aggregation time interval and the one or more auxiliary features;
inputting the plurality of timeseries datasets for the target feature into an anomaly detection model to identify one or more anomalies within the monitoring dataset; and
transmitting an alert message to the monitoring system, wherein the alert message comprises one or more indications of the one or more anomalies.
|