Purchase Limit Prediction Using Multi-Branch Regression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing neural network models for purchase prediction on users consider only a single factor, leading to inaccurate final purchase limits due to insufficient comprehensive consideration of relevant factors.
Innovation Solution
A multi-layer perceptron-based target model performs feature extraction, followed by multiple regression processes to obtain and combine purchase probabilities and limits from different perspectives, enhancing the accuracy of predicted purchase limits.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single regression processing is used to determine the final purchase limit, then the model complexity is reduced, but the prediction accuracy deteriorates due to insufficient comprehensive consideration of relevant factors
Solution Approach 1:
The patent divides the purchase limit prediction into multiple independent regression branches, each handling different aspects or types of purchase limit prediction. This segmentation allows each branch to specialize in specific factors, improving overall accuracy while maintaining manageable complexity through modular architecture.
Solution Approach 2:
The patent transitions from a single-dimensional prediction approach to a multi-dimensional approach by introducing multiple regression branches that operate in parallel. Each branch adds a new dimension of analysis, considering different factors simultaneously, thereby enriching the prediction with comprehensive considerations without overwhelming complexity.
2Adaptability or versatility
If multiple regression processes are used to obtain purchase limits from different perspectives, then the comprehensive consideration of factors is improved, but the model complexity increases
Solution Approach 1:
The patent designs a multi-functional regression system where multiple regression branches share common input layers and processing mechanisms while providing specialized output for different purchase limit perspectives. This universality allows the system to handle diverse prediction requirements through a unified architectural framework, reducing overall complexity.
Solution Approach 2:
The patent merges the outputs of multiple regression branches through a consolidation mechanism that integrates predictions from different perspectives into a final comprehensive purchase limit prediction. This merging process combines the strengths of each branch while filtering out redundancies, achieving comprehensive consideration without proportionally increasing complexity.
Data Source
AI summary
Example purchase prediction methods and apparatus are described. One example method includes obtaining information associated with a user. Feature extraction processing is performed on the information to obtain a purchase probability of the user. First regression processing is performed on the information to obtain a first purchase limit of the user. Second regression processing is performed on the information to obtain a second purchase limit of the user, where the first regression processing and the second regression processing are different regression processing. A predicted purchase limit of the user is obtained based on the purchase probability, the first purchase limit, and the second purchase limit.