Classification System with Unknown Class Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Classifiers using supervised learning models often deteriorate in classification precision when encountering data not belonging to any defined class, leading to incorrect classifications.
Innovation Solution
A classification system comprising a score calculation unit, a determination unit, and a classification unit that calculates scores for predetermined classes, determines whether the input data belongs to any of these classes, and classifies it as an unknown class if it does not match any, thereby preventing wrong classifications and maintaining precision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a classifier uses supervised learning models to classify data into predefined classes, then classification can be performed efficiently, but classification precision deteriorates when data not belonging to any defined class is input
Solution Approach 1:
The classification process is segmented into two distinct stages: first, the score calculation unit computes similarity scores between input data and each predefined class; second, the determination unit evaluates whether the input data belongs to any predefined class by comparing scores against a threshold. This segmentation allows the system to maintain high classification efficiency for known classes while accurately identifying unknown classes, thereby resolving the contradiction between productivity and measurement precision.
2Adaptability or versatility
If the classifier forces all input data into predefined classes, then classification coverage is maximized, but wrong classifications occur for unknown data
Solution Approach 1:
The determination unit acts as an intermediary between the score calculation unit and the classification output. It receives similarity scores from the score calculation unit, compares them against a predetermined threshold, and determines whether the input data belongs to any predefined class. This intermediary mechanism enables the system to maintain broad classification coverage for known classes while reliably identifying unknown classes by preventing forced classification, thus resolving the contradiction between adaptability and reliability.
Data Source
AI summary
A classification system according to an embodiment includes a score calculation unit, a determination unit, and a classification unit. The score calculation unit calculates respective scores of predetermined classes from input data. The determination unit determines whether the input data belongs to anyone of the classes based on the respective scores of the classes, which are calculated by the score calculation unit. The classification unit determines which one of the classes the input data belongs to, based on the calculated scores when the determination unit determines that the input data belongs to anyone of the classes and determines that the input data belongs to an unknown class that is other than the classes when the determination unit determines that the input data does not belong the classes.