Dynamic Demand Transfer Estimation Using Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing online retailing systems fail to accurately predict customer buying decisions in real-time due to dynamic changes in sales drivers such as stock availability and promotional factors, leading to inefficiencies in demand transfer estimation.
Innovation Solution
A processor-implemented method and system using machine learning to generate primary and secondary data matrices from transaction data, training a multivariate model to predict the probability of sales for each product and estimate demand transfer between products based on real-time sales drivers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional online retailing systems use static prediction models, then the system complexity is low, but the accuracy of demand transfer estimation deteriorates due to inability to capture real-time dynamic changes in sales drivers
Solution Approach 1:
The patent implements dynamic demand transfer estimation by continuously updating prediction models with real-time data on sales drivers such as stock availability, promotional factors, and customer behavior patterns. The system transitions from static to dynamic modeling, allowing accuracy to improve while managing complexity through structured data processing pipelines and automated model training cycles.
Solution Approach 2:
The system incorporates feedback mechanisms where actual customer purchasing decisions and real-time transaction data are fed back into the machine learning models to refine future predictions. This continuous feedback loop enables the system to learn from actual demand transfer patterns and adjust its estimation accuracy accordingly, resolving the contradiction between precision and complexity.
2Measurement precision
If the system considers more sales drivers and real-time factors, then the accuracy of customer buying decision prediction improves, but the data processing complexity and computational requirements worsen
Solution Approach 1:
The patent segments the data processing into distinct matrices: a primary data matrix containing customer demographics and sales driver information, and a secondary data matrix containing transactional data and demand transfer records. This segmentation allows the system to handle multiple sales drivers systematically, processing each dimension separately before integrating them into the final prediction model, thereby managing complexity while maintaining comprehensive analysis.
Solution Approach 2:
The system transforms raw transaction data into structured multivariate data matrices, adding dimensional structure to the data representation. By organizing data into matrices with specific dimensions (customer attributes, product attributes, temporal dimensions, and interaction dimensions), the system can process complex relationships between multiple sales drivers more efficiently, improving prediction accuracy without proportionally increasing processing complexity.
3Adaptability or versatility
If the system performs real-time demand transfer estimation, then the responsiveness to customer behavior changes improves, but the computational time and processing speed requirements worsen
Solution Approach 1:
The system performs preliminary processing of historical data to pre-compute and store customer profiles, product attributes, and demand transfer patterns in structured matrices. This preliminary action prepares the data in advance, allowing the real-time estimation to focus only on computing predictions for current transactions rather than processing all data from scratch, thus improving real-time responsiveness while managing computational speed requirements.
Data Source
AI summary
During online shopping, customer buying decision varies based on conditions at the time of logging such as product availability, competitor price of the product, presence of promotion, delivery options such as number of days to deliver, availability of free delivery, and availability of pay on delivery and customer review ratings. Customer shifts from one product to other product based on the options available at real time and accordingly demand of a product is transferred to other product. The method and system disclosed provides dynamic demand transfer values that are specific to a customer for available options at the time of login and it provides the list of ideal products to be displayed at the time of customer login. The method utilizes a suitable data format to apply machine learning based approach for estimating DT, wherein training data for ML captures plurality of sales drivers affecting customer decision during online retailing.