Tensor-Based Graph Processing Model Deficiency Data Objects

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing predictive data analysis solutions face inefficiencies in training speed and accuracy, often requiring a tradeoff between the two, and struggle with operational load balancing in post-prediction systems.

Innovation Solution

The development of a tensor-based graph processing machine learning framework that generates a model deficiency data object using holistic graph links inferred by a graph representation machine learning model, enhancing predictive accuracy and training speed while improving operational load balancing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional predictive data analysis solutions are used, then operational simplicity is maintained, but training speed and predictive accuracy deteriorate

Engineering Contradiction:

Improvepredictive accuracyVSAvoidtraining speed

Core Design Contradiction:

Measurement precisionVSProductivity

Solution Approach 1:

The patent segments the graph processing into multiple tensor-based representations (adjacency tensors, feature tensors, edge attribute tensors) that can be processed independently and in parallel. This segmentation enables faster training by distributing computational load across multiple tensor operations while maintaining accurate graph structure representation for predictive analysis.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent transitions from traditional flat data structures to multi-dimensional tensor representations of graphs. By representing graph data in tensor form with additional dimensionalities (nodes, edges, features, layers), the system achieves both higher predictive accuracy through richer feature interactions and improved training speed via efficient tensor parallelization.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

2Measurement precision

If model complexity is increased to improve predictive accuracy, then measurement precision improves, but device complexity increases

Engineering Contradiction:

Improvepredictive accuracyVSAvoidmodel complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent changes the parameter representation from traditional scalar or vector forms to multi-dimensional tensors. This parameter transformation enables the model to capture complex graph relationships and feature interactions more effectively, improving predictive accuracy while the structured tensor format actually reduces computational complexity through optimized linear algebra operations.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20240013064A1Machine learning techniques using model deficiency data objects for tensor-based graph processing models

Publication Date: 2024.01.11 OPTUM INC

AI summary

Various embodiments of the present invention provide methods, apparatus, systems, computing devices, computing entities, and/or the like for generating a model deficiency data object for a tensor-based graph processing machine learning model. Certain embodiments of the present invention utilize systems, methods, and computer program products that generate a model deficiency data object for a tensor-based graph processing machine learning model using holistic graph links generated by utilizing a graph representation machine learning model.