Estimation Model Clustering Items Selection Behavior
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for modeling and learning selection behavior affected by cognitive bias become complex when dealing with multiple choice sets, as the number of parameters to be learned increases, making it difficult to analyze and predict consumer behavior effectively.
Innovation Solution
A training method that clusters items based on their attribute values, generates cluster attribute values, and trains an estimation model using these values to predict selection behavior, thereby simplifying the learning process and considering cognitive biases in choice sets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If modeling is performed for each choice set separately, then selection behavior can be accurately captured, but the number of parameters to be learned increases and learning becomes difficult
Solution Approach 1:
The patent merges multiple choice sets into a unified model by introducing a choice set attribute that captures common characteristics across different choice sets. Instead of learning separate parameters for each choice set, the model learns shared parameters while adjusting for choice set-specific effects through the choice set attribute, thereby reducing the total number of parameters while maintaining modeling accuracy.
Solution Approach 2:
The patent transforms the parameter structure by changing from choice set-specific parameters to general parameters modified by choice set attributes. This parameter transformation allows the model to generalize across choice sets while still capturing their unique characteristics, reducing parameter complexity without sacrificing precision.
2Loss of information
If multiple choice sets are processed independently, then detailed selection patterns can be analyzed, but computational complexity increases and learning efficiency decreases
Solution Approach 1:
The patent creates a universal model structure that can handle multiple choice sets simultaneously through the choice set attribute. This multi-functional approach allows the same model parameters to be applied across different choice sets, enabling the system to process multiple choice sets efficiently while capturing their specific characteristics through the attribute mechanism.
Solution Approach 2:
The patent segments the influence of choice sets by separating general selection behavior parameters from choice set-specific effects. The choice set attribute captures the specific characteristics of each choice set, while the base model learns general patterns, allowing efficient joint processing of multiple choice sets without losing detailed selection patterns.
3Measurement precision
If cluster attribute values are generated for all items, then selection behavior estimation accuracy improves, but computational resources and time increase
Solution Approach 1:
The patent performs preliminary clustering of items based on their attributes before the main selection behavior estimation process. By pre-computing cluster assignments and cluster attribute values, the system avoids repeated computational overhead during estimation, thereby improving accuracy through clustering while minimizing the time penalty through advance preparation.
Data Source
AI summary
A training method is provided. The training method includes clustering, by a processor, a plurality of items that each have an item attribute value, according to the item attribute value. The training method further includes generating, by the processor, for each item, a cluster attribute value corresponding to a cluster associated with the item. The training method also includes training, by the processor, an estimation model for estimating selection behavior of a target with respect to a choice set including two or more items, based on the cluster attribute value associated with each item included in the choice set, by using training data that includes a group of a choice set of items presented to the target and an item selected by the target from among the choice set.