Vehicle Estimation Device Using Logit Correction for Class Imbalance

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing machine learning models face accuracy issues due to class imbalances in the number of samples, leading to decreased estimation performance.

Innovation Solution

A method to correct logit values using a hyperparameter and prior probability calculated from a marginal distribution, enhancing the accuracy of the machine learning model by adding a correction term to the logit values.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If simple logit correction using sample counts is applied, then the correction process is simple, but the logit correction accuracy is low

Engineering Contradiction:

Improvecorrection process complexityVSAvoidlogit correction accuracy

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The patent changes the parameters used for logit correction from simple sample counts to prior probabilities calculated from marginal distributions. This involves computing P(y|x) using the machine learning model's output, then deriving prior probabilities that reflect true class distributions, thereby improving correction accuracy while maintaining computational feasibility

Inventive Principle:

Principle #35Parameter changes

Solution Approach 2:

The patent introduces an intermediary calculation step that uses the machine learning model's output to compute marginal distributions and prior probabilities. This intermediary layer transforms the raw model output into corrected logits by applying the calculated prior probabilities, bridging the gap between simple counting and accurate probability adjustment

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS20250245980A1Estimation device, estimation method, storage medium, vehicle control device, and vehicle

Publication Date: 2025.07.31 HONDA MOTOR CO LTD

AI summary

An estimation device includes a storage medium configured to store computer-readable instructions, and a processor connected to the storage medium, in which the processor executes the computer-readable instructions to acquire, with data as an input, a logit that at least a portion of the data corresponds to a class that represents a certain type by inputting target data to a machine learning model learned to output the logit, calculate a correction value for correcting the logit using an output of the machine learning model, correct the logit on the basis of the calculated correction value, and estimate a class to which at least a portion of the target data corresponds on the basis of the corrected logit.