Estimation Model for Price Optimization Using Relative Prices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing price optimization models often produce unrealistic prices and struggle to accurately estimate demand in scenarios where non-price factors, such as travel date, significantly influence demand.
Innovation Solution
A computer-readable storage medium with instructions to train an estimation model using training data that includes relative prices and other features, generating an estimation function with a regularization term to optimize prices and improve demand prediction, incorporating timing and other transaction-related factors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If existing price optimization models are used, then automated price determination is achieved, but unrealistic optimal prices are produced
Solution Approach 1:
The patent transforms the price input parameter from absolute price to relative price (log difference from mean price). This parameter transformation stabilizes the training data distribution and enables the model to produce realistic optimal prices that are comparable to current prices, resolving the issue of unrealistic price estimates while maintaining automated price determination
Solution Approach 2:
The patent introduces relative price as an intermediary representation between absolute price and demand. By using relative price (log price difference from mean) as the input feature, the model achieves better convergence and produces realistic optimal prices that maintain the benefit of automated price determination
2Extent of automation
If existing price optimization models are used, then automated pricing is implemented, but difficulty arises when demand strongly depends on non-price factors
Solution Approach 1:
The patent designs a universal estimation model that handles both price-sensitive and non-price-sensitive demand scenarios through the same relative price feature. The model universally processes different types of demand patterns (price elastic and inelastic) by transforming price to relative price, enabling automated pricing to work accurately across diverse product types including travel packages where demand depends on timing factors
3Manufacturing precision
If regularization term is added to objective function, then optimal price is constrained to be closer to standard price, but model flexibility is reduced
Solution Approach 1:
The patent applies partial regularization by using a moderate regularization parameter lambda that partially constrains optimal price toward standard price. This partial action is sufficient to prevent unrealistic price estimates while maintaining enough model flexibility to adapt to different demand patterns and product types
Data Source
AI summary
A non-transitory computer readable storage medium having instructions embodied therewith, the instructions executable by a processor or programmable circuitry to cause the processor or programmable circuitry to perform operations including obtaining training data including a sample value of one or more input features of an item and a sample value of an output feature representing demand for the item, and training, based on the training data, an estimation model that estimates a new value of the output feature for the item based on new values of the one or more input features. The one or more input features may include a relative price of the item relative to prices of a plurality of items.