Inference Prediction Graphs for Reliable Data Processing Policies

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing data processing systems face challenges in managing operational states due to unreliable inference model predictions and complex, stochastic data, leading to potential service interruptions and reduced quality.

Innovation Solution

Utilize a plurality of inference models to generate predictions, analyze them statistically to obtain an acyclic graph representing clusters of future states, and define policies based on this graph to manage operations proactively.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If inference model predictions are used to manage operational states, then proactive policy management is enabled, but reliability is reduced due to unpredictable and complex stochastic data

Engineering Contradiction:

Improveproactive policy managementVSAvoidprediction reliability

Core Design Contradiction:

Extent of automationVSReliability

Solution Approach 1:

The patent introduces an acyclic graph as an intermediary data structure between the inference model predictions and the policy management system. The graph transforms complex, stochastic prediction data into a structured representation with nodes (operational states) and edges (transitions), making the data manageable and reliable for policy definition while preserving the proactive automation capability

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The system implements feedback by continuously monitoring operational states, comparing actual system behavior against predicted future states represented in the acyclic graph, and automatically adjusting policies based on deviations. This feedback loop enhances reliability by validating predictions against real-world outcomes while maintaining automated policy management

Inventive Principle:

Principle #23Feedback

2Measurement precision

If multiple inference models are used to generate predictions, then prediction accuracy is improved, but device complexity increases

Engineering Contradiction:

Improveprediction accuracyVSAvoidsystem complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent merges multiple inference model predictions into a unified acyclic graph structure, where predictions from different models are integrated into a single coherent representation of future operational states. This consolidation maintains prediction accuracy while reducing system complexity by providing a unified interface for policy management

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The acyclic graph serves as a universal data structure that can accommodate predictions from multiple different inference models with varying complexities. This multi-functional approach allows the system to leverage diverse prediction sources without increasing operational complexity, as the graph structure handles the integration uniformly

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Manufacturing precision

If detailed statistical characterization of predictions is performed, then policy definition accuracy is improved, but loss of time increases due to complex data analysis

Engineering Contradiction:

Improvepolicy definition accuracyVSAvoidanalysis time

Core Design Contradiction:

Manufacturing precisionVSLoss of time

Solution Approach 1:

The patent segments the complex statistical analysis into discrete, manageable components represented as nodes and edges in the acyclic graph. Each node represents a specific operational state with its own statistical characteristics, and each edge represents a transition with defined probabilities. This segmentation enables precise policy definition while reducing analysis time by allowing parallel processing of different state transitions

Inventive Principle:

Principle #1Segmentation

4Ease of operation

If acyclic graph is used to represent future states, then ease of operation is improved for policy management, but device complexity increases due to additional data structures

Engineering Contradiction:

Improvepolicy management easeVSAvoiddata structure complexity

Core Design Contradiction:

Ease of operationVSDevice complexity

Solution Approach 1:

The patent creates a simplified copy or abstraction of the complex prediction data in the form of an acyclic graph. This graphical representation copies only the essential elements (operational states and transitions) needed for policy management, making the system easier to operate while the underlying complexity is managed by automated processes that generate and maintain the graph structure

Inventive Principle:

Principle #26Copying

Data Source

PatentUS20260037839A1Managing policy definition for data processing systems using representations of inference model predictions

Publication Date: 2026.02.05 DELL PROD LP

AI summary

Methods and systems for managing operation of a data processing system are disclosed. To manage the operation, a plurality of predictions using at least one inference model trained to predict future states for the data processing system over time may be obtained. A statistical characterization regarding agreement in the plurality of predictions may be obtained. The statistical characterization and the plurality of predictions may be used to obtain an acyclic graph for use in analyzing the predicted future states. Based on analysis of the acyclic graph, a policy defining the operation of the data processing system may be defined, and a computer-implemented service may be provided using the data processing system in accordance with the policy.