Reliability Determination Unit for Inference Accuracy Verification

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The inference accuracy of an electric device group's operation state deteriorates when training data associating feature values with operation states is not appropriately set, and it is difficult to check if the training data is correctly set.

Innovation Solution

An information providing apparatus and method that includes a reliability determination unit to assess the reliability of inference results based on time-series measurement data, such as current or power consumption, and training data associating operation states with feature values, and an output unit to provide information on the reliability of the inference results.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If training data is not appropriately set, then the inference system can operate without additional verification mechanisms, but the inference accuracy of the electric device group's operation state deteriorates and it becomes difficult to check if training data is correctly set

Engineering Contradiction:
Improveinference accuracyVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent implements a feedback mechanism by calculating a reliability index that compares the inferred operation state with the actual operation state of the electric device group. This reliability index is fed back to assess whether the training data is appropriately set, enabling continuous verification and improvement of inference accuracy without requiring complex manual inspection procedures

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The system performs self-verification by automatically calculating the reliability index using the inference results and actual operation states. This self-service mechanism allows the system to autonomously check whether training data is correctly set without requiring external verification, thereby improving reliability while maintaining simple operation

Inventive Principle:
Principle #25Self-service

2Measurement precision

If training data is not appropriately set, then the system structure remains simple, but the inference accuracy deteriorates and verification becomes difficult

Engineering Contradiction:
Improveinference accuracyVSAvoidease of verification
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The patent introduces a reliability index as an intermediary metric that bridges the gap between inference results and training data quality assessment. This intermediary enables easy verification of training data appropriateness by providing a quantifiable measure that can be straightforwardly calculated and interpreted, making the verification process simple while maintaining high inference accuracy

Inventive Principle:
Principle #24Intermediary (Mediator)

3Reliability

If reliability determination is added to the inference system, then the reliability of inference results improves, but the device complexity increases

Engineering Contradiction:
Improvereliability of inference resultVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent extracts the essential verification function by focusing solely on calculating a reliability index based on the comparison between inferred and actual operation states. This extraction approach adds minimal complexity by implementing only the necessary reliability determination component rather than a comprehensive verification system, thereby improving reliability while maintaining system simplicity

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentUS10750251B2Information providing apparatus, information providing method, and storage medium
Publication Date: 2020.08.18 NEC CORP
  • US10750251B2 patent drawing
  • US10750251B2 patent drawing
  • US10750251B2 patent drawing

AI summary

Provided is an information providing apparatus (10) including a reliability determination unit (11) that determines a reliability of an inference result obtained by inferring an operation state of a monitored electric device group, on the basis of time-series measurement data and training data, the time-series measurement data including any one of current consumption and power consumption of the monitored electric device group and the training data associating each of a plurality of the operation states which can be assumed by the electric device group with a feature value of each of the plurality of operation states, and an output unit (12) that outputs information related to the reliability.