Machine Learning Segmentation With Interactive Cluster Refinement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing segmentation methods in supply chain management are time-consuming, expensive, and require significant human resources due to the complexity of managing numerous items with diverse attributes, making it difficult to group similar items efficiently.
Innovation Solution
An automated machine learning segmentation tool that quickly segments items by receiving attributes, engineering features, training cluster-based models, and allowing user interaction for iterative refinement, reducing time and resource requirements while enhancing accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional segmentation methods are used to group items by common attributes, then items can be managed in manageable groups, but the process requires months to years, significant human resources, and specialized experts
Solution Approach 1:
The patent replaces the manual, expert-driven mechanical process of segmentation with an automated machine learning system. The ML model automatically processes item attributes, performs feature engineering, and generates segments without requiring specialized human experts, thereby dramatically reducing both time and resource requirements while maintaining or improving segment quality.
Solution Approach 2:
The segmentation system performs self-service by automatically conducting feature engineering, model training, and segment generation without continuous human intervention. The system autonomously processes the segmentation workflow from raw attributes to final segments, enabling rapid processing of millions of items in minutes rather than months or years.
2Reliability
If manual segmentation is performed with consolidated data and planner judgement, then meaningful groups can be created, but the process becomes very expensive due to specialized experts and large teams
Solution Approach 1:
The patent substitutes the complex human expert system with an automated machine learning system. The ML model replicates and enhances the judgment capability of planners by learning from historical data and automatically identifying meaningful patterns, thereby maintaining segmentation quality while eliminating the need for expensive specialized experts and large teams.
Solution Approach 2:
The system changes the parameters of the segmentation process by using automated feature engineering and ML model training instead of manual analysis. This transformation converts the complex human-dependent process into a scalable computational process that maintains reliability while reducing resource complexity.
3Productivity
If automated machine learning segmentation is used to process millions of items quickly, then segmentation speed increases to minutes, but the system requires sophisticated feature engineering and model training
Solution Approach 1:
The patent applies segmentation at the feature level by automatically engineering features from raw attributes and then segmenting items based on these engineered features. This multi-level segmentation approach (feature segmentation followed by item segmentation) enables the system to handle complexity internally while maintaining high throughput and rapid processing of millions of items.
Solution Approach 2:
The system performs preliminary feature engineering and model training actions automatically before the actual segmentation occurs. By pre-processing the data and training models in advance, the system prepares the computational infrastructure to rapidly segment millions of items in minutes without requiring complex manual intervention during the segmentation execution phase.
Data Source
AI summary
Machine learning segmentation methods and systems that perform segmentation quickly, efficiently, cheaply, and optionally provides an interactive feature that allows a user to alter the segmentation until a desired result is obtained. The automated machine learning segmentation tool receives all potentially important attributes and provides segmentation of items. It also receives information about important features of the data and finds how best to differentiate between groups using cluster-based machine learning algorithms. In addition, visualization of the segmentation explains to a user how the segmentation was obtained.