Extrapolated Indication Generation Using Statistical Data Brackets

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing models face challenges in accurately predicting subsequent events with high fidelity and efficiency due to the need for large amounts of data, which increases processing time and reduces accuracy when reducing redundancy in model input data.

Innovation Solution

A system utilizing multiple variable statistical algorithms and decision tree analysis to identify and generate determined variables and values that correlate strongly with previous events, focusing on data brackets that define strong interpolated correlations to generate an extrapolated indication for subsequent events, thereby reducing unnecessary data processing and improving accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a large amount of model data is provided to determine parameters, then the accuracy of the inference with respect to the subsequent outcome increases, but the processing time and computing power consumption increase

Engineering Contradiction:

Improveaccuracy of inferenceVSAvoidprocessing time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent extracts and identifies only the most relevant parameters from the large amount of model data using statistical algorithms. By determining which parameters have the strongest correlation with subsequent event indications, the system extracts a reduced set of essential parameters that maintain inference accuracy while significantly reducing processing requirements.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent transforms the approach by changing from using all available parameters to using a selected subset of parameters. Through statistical analysis, the system identifies and switches to using only the most influential parameters, thereby changing the parameter quantity from large to reduced while maintaining or improving efficiency.

Inventive Principle:

Principle #35Parameter changes

2Measurement precision

If a large amount of model data is provided to determine parameters, then the accuracy of the inference with respect to the subsequent outcome increases, but the computing power consumption increases

Engineering Contradiction:

Improveaccuracy of inferenceVSAvoidcomputing power consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

Inventive Principle:

Principle #35Parameter changes

3Productivity

If the model data is reduced to improve processing efficiency, then the processing time and computing power consumption decrease, but the accuracy of the characteristic determination with respect to the subsequent outcome reduces

Engineering Contradiction:

Improveprocessing efficiencyVSAvoidaccuracy of characteristic determination

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent employs statistical algorithms to analyze the relationship between parameters and subsequent event indications, using feedback from the data to identify which parameters are most predictive. This feedback mechanism ensures that the reduced parameter set maintains high accuracy by continuously validating which parameters provide the most useful information.

Inventive Principle:

Principle #23Feedback

4Quantity of substance

If redundant data is included in the model input, then the quantity of data increases, but the fidelity of the extrapolated indication and efficiency decrease

Engineering Contradiction:

Improvequantity of model dataVSAvoidfidelity of extrapolated indication

Core Design Contradiction:

Quantity of substanceVSReliability

Solution Approach 1:

Inventive Principle:

Principle #2Taking out (Extraction)

Data Source

PatentUS20240193437A1Efficient generation of an extrapolated indication associated with a subsequent event without redundancy

Publication Date: 2024.06.13 TRUIST BANK

AI summary

A system with a processor configured to perform steps including receive determination data indicative of previous events associated with users and to identify, utilizing a multiple variable statistical algorithm, data brackets, each having interpolated correlation with indications of the previous events. Further steps include to identify, utilizing bivariate analysis and the data brackets, a determined variable defining a strong interpolated correlation with indications of the previous events and to generate, utilizing bivariate analysis, an associated determined value that separates ranges of data associated with the determined variable based on the strong interpolated correlation within the determination data. Furthermore, the identified determined variable and generated determined value, when utilized to generate an extrapolated indication associated with a subsequent event, rather than input data of a data bracket not indicating the determined variable for the subsequent event, increases the fidelity of the extrapolated indication, reduces redundancy within the extrapolated indication, or both.