Guided Hierarchical Classification for Minority Event Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Standard data classification approaches often fail to effectively handle minority events, which can occur infrequently but result in significant misclassification costs, particularly in applications like medical diagnosis.
Innovation Solution
A guided hierarchical classification algorithm using a decision tree structure is employed to isolate minority classes by iteratively distributing classes into child nodes, allowing for precise classification and reducing misclassification errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If standard classification approaches are used, then the system is simple and fast, but minority events are misclassified with high error rates
Solution Approach 1:
The patent divides the classification task into multiple stages using an ensemble of classification algorithms. Different algorithms are applied in sequence or parallel to progressively refine the classification of minority events, breaking down the complex problem into manageable segments that can be handled by individual algorithms with specialized functions.
Solution Approach 2:
The patent employs an ensemble approach where multiple classification algorithms with different strengths are combined into a single system. This multi-functional ensemble can handle both majority and minority classes effectively, with each algorithm contributing its unique capability to the overall classification task.
2Reliability
If standard classification approaches are used, then training is fast and simple, but detection of rare events is unreliable
Solution Approach 1:
The patent performs preliminary actions during the training phase by pre-training multiple classification algorithms and establishing their ensemble configuration beforehand. This preliminary preparation enables the system to reliably detect minority events during deployment without requiring extensive processing time for each individual classification task.
Solution Approach 2:
The patent creates multiple copies of classification algorithms with different configurations and training data subsets. These copied algorithms are then combined in an ensemble, allowing the system to leverage diverse perspectives and patterns for more reliable minority event detection while distributing the training workload across multiple parallel processes.
Data Source
AI summary
Technologies for detecting minority events are disclosed. By performing a guided hierarchical classification algorithm with a decision tree structure and grouping the minority class(es) in with some of the majority classes, large majority classes may be separated from a minority class without requiring good detection of the minority events by themselves. The decision tree structure may be used only for the purpose of identifying if the data sample in question is a member of a minority class. If it is determined that it is not, a primary classification algorithm may be used. With this approach, the guided hierarchical classification algorithm need not perform as well as the primary classification algorithm for the majority events, but may provide improved detection for minority events.