AI Anomaly Detection Model Training from Normal Time Intervals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional log anomaly detection approaches require manual filtering of training data and are time- and resource-intensive, leading to errors.
Innovation Solution
Automatically training and implementing artificial intelligence-based anomaly detection models by analyzing observability data, identifying normal intervals, and using clustering algorithms to filter out abnormal signals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual filtering of training data is used in conventional anomaly detection approaches, then data can be processed, but the process becomes time- and resource-intensive and error-prone
Solution Approach 1:
The system performs self-training by automatically identifying normal intervals from time series data without requiring manual filtering of training data. The anomaly detection model trains itself using the identified normal intervals as training samples, eliminating the need for human intervention in data preparation while maintaining detection accuracy.
Solution Approach 2:
The system performs preliminary identification of normal intervals from time series data before the actual anomaly detection process. This preliminary action automatically prepares the training data by identifying periods with given levels of data activity, so that when anomalies need to be detected, the model is already trained and ready, saving time.
2Reliability
If manual filtering of training data is used in conventional anomaly detection approaches, then data can be processed, but resource consumption increases
Solution Approach 1:
The system automatically identifies normal intervals and trains the anomaly detection model without requiring manual data filtering processes. This self-service approach reduces the computational overhead associated with manual data preparation and filtering, while maintaining the reliability of anomaly detection through automated model training.
3Productivity
If automated anomaly detection is implemented, then time and resources are saved, but the system must automatically identify normal intervals from time series data
Solution Approach 1:
The system segments the time series data into distinct intervals based on data activity levels, automatically identifying normal intervals for training. This segmentation approach breaks down the complex task of anomaly detection into manageable steps: identifying normal periods, training the model on these periods, and then detecting anomalies, thereby reducing overall system complexity while improving productivity.
Data Source
AI summary
Methods, systems, and computer program products for automatically detecting periods of normal activity by analyzing observability data in IT operations environments are provided herein. A computer-implemented method includes obtaining multiple types of data related to one or more artificial intelligence-related information technology operations; modelling at least a portion of the obtained data as time series data; automatically identifying, from the time series data, one or more time periods associated with one or more given levels of data activity; and performing 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.