Anomaly Detection Using Pairwise Interaction Score Models

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current predictive data analysis solutions are inefficient and unreliable due to resource-intensive and time-consuming manual processes for detecting event code anomalies in event data objects, leading to errors and suboptimal resource allocation.

Innovation Solution

The method involves generating a target event data object with event codes weighted based on primary and related event data objects, using a pairwise interaction score determination machine learning model to refine event code pairs, and creating anomaly detection matrices to identify an anomalous code subset, thereby automating anomaly detection and optimizing resource allocation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If manual processes are used for detecting event code anomalies, then detection can be performed, but the process is resource-intensive and time-consuming

Engineering Contradiction:

Improveanomaly detection reliabilityVSAvoiddetection time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent replaces manual mechanical review processes with an automated machine learning system that uses trained models to detect anomalous event codes. The system automatically processes event data objects, generates anomaly detection matrices, and identifies anomalies without human intervention, thereby eliminating the time-consuming nature of manual processes while maintaining or improving detection reliability.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables self-service anomaly detection by automatically processing event data through trained machine learning models. The anomaly detection model autonomously analyzes event codes, generates detection matrices, and identifies anomalies without requiring manual analysis, allowing the system to serve itself in detecting and flagging anomalous patterns efficiently.

Inventive Principle:

Principle #25Self-service

2Reliability

If manual processes are used for detecting event code anomalies, then detection can be performed, but computational resource expenditure is high

Engineering Contradiction:

Improveanomaly detection reliabilityVSAvoidcomputational resource expenditure

Core Design Contradiction:

ReliabilityVSUse of energy by moving object

Solution Approach 1:

The patent applies preliminary action by training machine learning models in advance on historical event data before deployment. The models are pre-trained to recognize normal and anomalous patterns, so during actual operation, they can quickly process new event data without requiring intensive real-time computational resources. This upfront preparation reduces the computational burden during production use.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system uses partial action by focusing computational resources only on evaluating event codes against the pre-trained anomaly detection model, rather than performing exhaustive manual analysis of all event data. The model selectively identifies potentially anomalous codes for further review, reducing overall computational expenditure while maintaining detection reliability.

Inventive Principle:

Principle #16Partial or excessive action

3Reliability

If manual review processes are used, then anomaly detection can be performed, but errors occur and resource allocation is suboptimal

Engineering Contradiction:

Improvedetection accuracyVSAvoidresource allocation efficiency

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The patent replaces error-prone manual review processes with automated machine learning models that consistently apply detection criteria without human error. The system automatically processes event data, generates anomaly detection matrices, and identifies anomalies with high accuracy, eliminating the errors inherent in manual review while optimizing resource allocation through automated workflow management.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system incorporates feedback mechanisms where the anomaly detection model continuously learns from detected anomalies and adjusts its detection parameters. The model receives feedback from confirmed anomalies and refines its detection accuracy over time, improving reliability while maintaining efficient resource allocation through automated iterative improvement.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS11914506B2Machine learning techniques for performing predictive anomaly detection

Publication Date: 2024.02.27 OPTUM INC

AI summary

Various embodiments of the present invention provide methods, apparatus, systems, computing devices, computing entities, and/or the like for performing predictive data analysis operations. For example, certain embodiments of the present invention utilize systems, methods, and computer program products that perform predictive data analysis operations by determining an anomalous event code subset using a pairwise interaction score determination machine learning model.