Error Determination Apparatus for Classification Accuracy and Probability Output

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional data classification techniques can determine the accuracy of classification but fail to output probabilities indicating the likelihood of data belonging to specific classes.

Innovation Solution

An error determination device comprising a classification estimation process observation unit, a probability estimation unit, and an error determination unit that generates an estimation process feature vector, calculates estimated probabilities, and determines the correctness of classification results, outputting both the classification result and the estimated probability vector.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional data classification techniques are used, then classification accuracy determination is achieved, but probability output for each class is not provided

Engineering Contradiction:
Improveclassification accuracy determinationVSAvoidprobability information
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent segments the classification output into two distinct components: (1) a determination result indicating whether classification is correct or incorrect, and (2) a probability vector containing probability values for each class. This segmentation allows the system to provide both accuracy determination and probability information separately, resolving the contradiction between achieving accurate classification determination and providing complete probability information.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent adds a new dimension to the classification output by introducing the probability vector as a separate output component. Instead of only providing a binary correct/incorrect determination, the system now operates in an enhanced output space that includes both the determination result and the probability distribution across classes, thereby preventing loss of probability information.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Productivity

If conventional classification techniques are used, then processing speed is maintained, but additional probability estimation computation is required

Engineering Contradiction:
Improveclassification processing speedVSAvoidcomputation complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent merges the probability estimation computation with the existing classification determination process. By integrating the probability vector generation into the same computational framework used for accuracy determination, the system avoids separate independent computations, thereby managing computation complexity while providing both determination results and probability information.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The classification system is designed to perform multiple functions simultaneously: it determines whether classification is correct or incorrect and generates the probability vector for each class. This multi-functionality allows the same computational process to serve both accuracy assessment and probability estimation purposes, optimizing processing efficiency.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentUS20240311691A1Error determination apparatus, error determination method and program
Publication Date: 2024.09.19 NT T INC
  • US20240311691A1 patent drawing
  • US20240311691A1 patent drawing
  • US20240311691A1 patent drawing

AI summary

An error determination device comprising: a classification estimation process observation unit that acquires data in an estimation process from a classification estimation unit that estimates classification of data to be classified, and generates an estimation process feature vector on a basis of the data; a probability estimation unit that generates an estimated probability vector including probabilities each of which is a probability that the data to be classified belongs to one of classes on a basis of the estimation process feature vector; and an error determination unit that determines whether a classification result by the classification estimation unit is correct or incorrect on a basis of the estimated probability vector, and outputs the classification result, a determination result as to whether the classification result is correct or incorrect, and the estimated probability vector.