Aggregate Logit Sourcing Engine for Consumer Choice Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current choice modeling techniques, such as the multinomial logit model, assume all products are equally substitutable, leading to inaccuracies when modeling consumer choices among products that are not similar, and require significant computational time for complex sourcing scenarios.
Innovation Solution
The implementation of an aggregate logit sourcing engine that generates a closed-form model by creating multiple copies of a multinomial logit model, each with item utility parameters, and uses a matrix structure to calculate choice probabilities, minimizing the independence of irrelevant alternatives property and reducing computational burden.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional multinomial logit model is used to model consumer choices, then the model is simple to implement, but it assumes all products are equally substitutable leading to inaccuracies for diverse product sets
Solution Approach 1:
The patent segments the product set into multiple sourcing groups based on product similarity and substitutability. Each group is modeled separately with its own utility parameters, allowing the model to capture different substitution patterns for different product categories while maintaining computational tractability through the closed-form aggregation formula.
Solution Approach 2:
The patent creates a composite choice model that combines multiple multinomial logit models (one for each sourcing group) into a single aggregate model. The closed-form solution aggregates the choice probabilities from all groups using a weighted combination formula, producing a composite model that accurately represents diverse product substitutability patterns without requiring complex simulation methods.
2Measurement precision
If complex sourcing scenarios are modeled with traditional techniques, then more accurate consumer behavior representation is achieved, but significant computational time is required
Solution Approach 1:
The patent performs preliminary segmentation of products into sourcing groups before the choice modeling process. By pre-organizing products based on substitutability relationships and pre-calculating the aggregation weights for each group, the model avoids complex iterative computations during the actual choice probability calculation, achieving both accuracy and computational efficiency.
Solution Approach 2:
The patent creates simplified copy models for each sourcing group that replicate the essential substitution patterns within that group. Each copy model is a standard multinomial logit model that can be quickly evaluated, and the results are aggregated using a closed-form formula, avoiding the need for computationally intensive simulation while preserving accuracy.
3Adaptability or versatility
If analyst discretion is used to model product substitutability, then flexibility in modeling is achieved, but statistical repeatability and objectivity are reduced
Solution Approach 1:
The patent enables the model to automatically determine product substitutability relationships and sourcing group assignments based on observed choice data. The system self-organizes products into sourcing groups and calculates aggregation weights directly from the data without requiring analyst intervention, ensuring both flexibility in capturing actual substitution patterns and statistical repeatability through objective, data-driven parameter estimation.
Data Source
AI summary
Methods and apparatus are disclosed to model consumer choices. An example method includes adding, with a processor, a set of products having respondent choice data to a base multinomial logit (MNL) model, the base MNL model including an item utility parameter and a price utility parameter associated with corresponding ones of products in the set of products, generating, with the processor, a number of copies of the base MNL model to form an aggregate model based on a number of the corresponding ones of products in the set of products, each one of the number of copies of the base MNL model exhibiting an effect of an independence of irrelevant alternatives (IIA) property, proportionally affecting interrelationships, with the processor, between dissimilar ones of the number of products in the set by inserting sourcing effect values in the aggregate model to be subtracted from respective ones of the item utility parameters, estimating, with the processor, the item utility parameters of the aggregate model based on the number of copies of the base MNL model and the respondent choice data, and calculating, with the processor, the choice probability for the corresponding ones of the products in the set of products based on the estimated item utility parameters and the price utility parameters.