Automatically training and implementing artificial intelligence-based anomaly detection models

A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computing device to cause the computing device to:

• obtain multiple types of data related to one or more artificial intelligence-related information technology operations;
• model at least a portion of the obtained data as time series data;
• generate, using the time series data, a candidate set of time periods associated with one or more expected levels of data activity;
• extract features from the multiple types of data related to one or more artificial intelligence-related information technology operations;
• generate, using one or more of feature pooling techniques and at least one count vector-based recurrent neural network, a representation for each of at least a portion of time periods in the candidate set of time periods in connection with aggregating at least a portion of the extracted features;
• automatically identify, from the candidate set of time periods and based at least in part on the generated representation for each of the at least a portion of time periods in the candidate set of time periods, a subset of one or more time periods associated with the one or more expected levels of data activity, wherein automatically identifying the one or more time periods comprises automatically determining, using at least one sliding window technique, one or more data window sizes related to the one or more expected levels of data activity;
• and
• perform one or more automated actions, in at least one artificial intelligence-related information technology operations environment, based at least in part on the data corresponding to the one or more identified time periods, wherein performing one or more automated actions comprises:
• training at least one artificial intelligence-based anomaly detection model using the data corresponding to the one or more identified time periods;
• detecting one or more anomalies in one or more data sources associated with the at least one artificial intelligence-related information technology operations environment by processing data from the one or more data sources using the at least one trained artificial intelligence-based anomaly detection model;
• and
• initiating, based at least in part on the one or more detected anomalies, at least one automated action in connection with at least one system associated with the at least one artificial intelligence-related information technology operations environment.

A system comprising:

• a memory configured to store program instructions;
• and
• a processor operatively coupled to the memory to execute the program instructions to:
• obtain multiple types of data related to one or more artificial intelligence-related information technology operations;
• model at least a portion of the obtained data as time series data;
• generate, using the time series data, a candidate set of time periods associated with one or more expected levels of data activity;
• extract features from the multiple types of data related to one or more artificial intelligence-related information technology operations;
• generate, using one or more of feature pooling techniques and at least one count vector-based recurrent neural network, a representation for each of at least a portion of time periods in the candidate set of time periods in connection with aggregating at least a portion of the extracted features;
• automatically identify, from the candidate set of time periods and based at least in part on the generated representation for each of the at least a portion of time periods in the candidate set of time periods, a subset of one or more time periods associated with the one or more expected levels of data activity, wherein automatically identifying the one or more time periods comprises automatically determining, using at least one sliding window technique, one or more data window sizes related to the one or more expected levels of data activity;
• and
• perform one or more automated actions, in at least one artificial intelligence-related information technology operations environment, based at least in part on the data corresponding to the one or more identified time periods, wherein performing one or more automated actions comprises:
• training at least one artificial intelligence-based anomaly detection model using the data corresponding to the one or more identified time periods;
• detecting one or more anomalies in one or more data sources associated with the at least one artificial intelligence-related information technology operations environment by processing data from the one or more data sources using the at least one trained artificial intelligence-based anomaly detection model;
• and
• initiating, based at least in part on the one or more detected anomalies, at least one automated action in connection with at least one system associated with the at least one artificial intelligence-related information technology operations environment.