Telefonaktiebolaget lm ericsson (publ) (20240289209). METHOD AND APPARATUS FOR DETECTING AND EXPLAINING ANOMALIES simplified abstract

METHOD AND APPARATUS FOR DETECTING AND EXPLAINING ANOMALIES

Organization Name

telefonaktiebolaget lm ericsson (publ)

Inventor(s)

Razieh Abbasi Ghalehtaki of Montreal, Québec (CA)

Fetahi Wuhib of Pincourt, Québec (CA)

Amin Ebrahimzadeh of LaSalle, Québec (CA)

Roch Glitho of Ville Saint Laurent, Québec (CA)

METHOD AND APPARATUS FOR DETECTING AND EXPLAINING ANOMALIES - A simplified explanation of the abstract

This abstract first appeared for US patent application 20240289209 titled 'METHOD AND APPARATUS FOR DETECTING AND EXPLAINING ANOMALIES

The abstract describes a method and apparatus for detecting and explaining anomalies in data obtained from an environment using an encoder-decoder machine learning model. The model is trained with a first set of data samples, each comprising values for a plurality of features representing the state of the environment.

• The method involves determining thresholds for each feature based on maximum reconstruction errors found during training.
• An anomalous data sample is obtained and reconstruction errors for each feature in the sample are determined using the trained model.
• Features responsible for the anomaly are identified based on reconstruction errors exceeding the respective thresholds.

Potential Applications: - Anomaly detection in various industries such as manufacturing, healthcare, and finance. - Quality control in production processes. - Predictive maintenance in equipment monitoring.

Problems Solved: - Efficiently detecting anomalies in complex data sets. - Providing explanations for anomalies to aid in problem-solving.

Benefits: - Improved data analysis and decision-making. - Early detection of issues leading to cost savings. - Enhanced system reliability and performance.

Commercial Applications: Title: "Anomaly Detection System for Industrial Processes" This technology can be used in manufacturing plants, hospitals, and financial institutions to enhance operational efficiency, reduce downtime, and improve overall productivity.

Questions about Anomaly Detection System: 1. How does the encoder-decoder machine learning model help in detecting anomalies in data? The model is trained to reconstruct input data, and anomalies are identified based on reconstruction errors exceeding predefined thresholds.

2. What are the potential benefits of using this system in industrial processes? The system can lead to cost savings, improved quality control, and enhanced predictive maintenance strategies.

Original Abstract Submitted

embodiments described herein relate to a method and apparatus for detecting and explaining anomalies in data obtained from an environment using an encoder-decoder machine learning model. a state of the environment is represented by a plurality of features, and the machine learning model is trained with a first set of data samples. each data sample in the first set of data samples comprises values for each of the plurality of features. the method comprises determining a respective first threshold for each of the plurality of features based on respective maximum reconstruction errors for each feature found during training of the encoder-decoder machine learning model; obtaining an anomalous data sample; determining respective reconstruction errors for each feature in the anomalous data sample using the trained encoder-decoder machine learning model; and determining one or more features in the anomalous data sample that are responsible for the anomalous data sample being anomalous responsive to the reconstruction errors associated with the one or more features being greater than or equal to the respective first thresholds.