Encoded State-Change Clustering for Outcome Search in Multidimensional Systems

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Analyzing past system state instances in complex, multi-dimensional stateful systems to identify relationships with future system states is computationally overwhelming due to the volume of data and fleeting changes, making traditional analysis impractical.

Innovation Solution

Cluster unique machine-learnt encoded historical system state changes based on probable relationships with subsequent outcomes, using an LSTM autoencoder to reduce data volume and facilitate efficient search and analysis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional analysis methods are used to review past system state instances, then comprehensive analysis coverage is achieved, but computational burden becomes overwhelming and analysis becomes impractical

Engineering Contradiction:
Improveanalysis coverageVSAvoidcomputational efficiency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent extracts only the relevant and distinctive features from historical system state instances using machine learning encoding. Instead of analyzing complete high-dimensional state data, the system identifies and extracts key state change characteristics that are most likely to be related to future outcomes, thereby reducing computational burden while maintaining analysis effectiveness

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent segments the large volume of historical system state data into smaller, manageable clusters based on similarity of state changes. By dividing the comprehensive dataset into representative clusters, the system achieves practical analysis through sampling while preserving the essential patterns that would exist in complete analysis

Inventive Principle:
Principle #1Segmentation

2Loss of information

If complete historical system state data is reviewed, then all potential relationships are captured, but the volume of data makes analysis impractical

Engineering Contradiction:
Improverelationship detection completenessVSAvoiddata volume
Core Design Contradiction:
Loss of informationVSQuantity of substance

Solution Approach 1:

The patent performs preliminary machine learning encoding and clustering of historical system state data before actual analysis. By pre-processing and organizing the data into structured clusters with extracted features, the system prepares the data in advance to enable efficient relationship detection without requiring complete review of all raw historical data

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent transforms the high-dimensional system state data into encoded feature representations through machine learning. This parameter transformation converts complex state instances into compressed feature vectors that capture essential relationships while dramatically reducing data volume for analysis

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4660919A1Facilitation of analysis of past, or prediction of future, occurrences of a particular state change of a multidimensional system via clustering of machine learnt encoded historical system state changes based on a probable relationship with a subsequent system state change
Publication Date: 2025.12.10 CHICAGO MERCANTILE EXCHANGE INC
  • EP4660919A1 patent drawingFigure 1
  • EP4660919A1 patent drawingFigure 2
  • EP4660919A1 patent drawingFigure 3

AI summary

The disclosed embodiments relate to reducing a computational burden for identifying and analyzing past system state instances of complex, i.e., multidimensional or multi-variate, stateful systems which process large volumes of arbitrary or pseudo arbitrary transactions which modify the state thereof, where a prior system state instance may have an effect on a future system state instance, in order to, for example, discern some insight about actual or potential later occurring state instances. The disclosed embodiments cluster unique machine learnt encoded historical system state changes based on a probable/predictive relationship with one or more defined outcomes, each comprising one or more subsequent system state changes indicative thereof, forming an efficient outcome searchable database.