Multi-Polytope Classifier Boundaries for Overlapping Data Classes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional classification methods, such as support vector machines (SVM), face inefficiencies due to undesirable overlap between positive and negative data point classes, leading to inaccurate classification and increased computational resources.
Innovation Solution
A method integrating semi-supervised k-means clustering and SVM to construct multi-polytopes for data points, using separation hyperplanes to separate clusters, allowing for non-linear boundaries and improved accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a single hyperplane is used for classification, then the device complexity is low, but the classification accuracy deteriorates due to undesirable overlap between classes
Solution Approach 1:
The patent divides the single hyperplane classifier into multiple hyperplanes that form polytope structures. Each polytope encloses a specific class of data points, creating segmented classification regions. This segmentation allows for better separation of overlapping classes while maintaining a structured approach to complexity management.
Solution Approach 2:
The patent transitions from single-hyperplane (linear) classification to multi-hyperplane polytope classification, effectively adding dimensional complexity to the decision boundary. By constructing polytopes with multiple facets in the feature space, the classifier can capture non-linear relationships and complex class distributions that a single hyperplane cannot represent.
2Reliability
If conventional SVM with single hyperplane is used, then the computational resources are low, but the classification accuracy deteriorates
Solution Approach 1:
The computational task is segmented into multiple stages: first clustering data points into groups, then constructing polytopes for each cluster, and finally creating hyperplanes for separation. This segmentation allows the algorithm to focus computational effort on relevant data subsets rather than processing all data points uniformly, improving efficiency while enhancing accuracy.
Solution Approach 2:
The patent performs preliminary clustering of data points before constructing the final classifier. By pre-organizing data into clusters that correspond to different classes or sub-classes, the subsequent polytope construction and hyperplane generation become more efficient and accurate, as the algorithm starts with structured data rather than raw unprocessed points.
3Reliability
If multiple polytopes are constructed for clustered classes, then the classification accuracy is improved, but the device complexity increases
Solution Approach 1:
The classifier is segmented into modular polytope units, each handling a specific cluster of data points. This modular structure allows for systematic management of complexity, where each polytope can be independently constructed and optimized. The overall system complexity is managed by organizing these modular units in a hierarchical manner based on clustering results.
Data Source
AI summary
A computer implemented method of generating a classifier engine for machine learning includes receiving a set of data points. A semi-supervised k-means process is applied to the set of data points from each class. The set of data points in a class is clustered into multiple clusters of data points, using the semi-supervised k-means process. Multi-polytopes are constructed for one or more of the clusters from all classes. A support vector machine (SVM) process is run on every pair of clusters from all classes. Separation hyperplanes are determined for the clustered classes. Labels are determined for each cluster based on the separation by hyperplanes.