Quote Variable Prediction Using Dimension Reduction and Lookup Tables
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current configure, price, and quote applications face computational inefficiencies in generating real-time recommended value ranges for quote variables due to the combinatorial explosion of possible input combinations, making it costly and resource-intensive to calculate the highest-density interval (HDI) for probability density functions (PDFs).
Innovation Solution
The system employs dimension reduction techniques like Principal Component Analysis (PCA) to reduce the number of input combinations, groups related inputs, precomputes PDFs, and stores them as discrete points in a look-up table, enabling efficient real-time HDI calculations by identifying applicable PDFs during the quote-creation process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the system calculates HDI for each possible combination of input variables and values to provide accurate recommended price ranges, then the prediction accuracy and confidence indication are improved, but the computational cost and resource requirements increase exponentially due to combinatorial explosion
Solution Approach 1:
The system pre-calculates and stores HDI values for all possible input combinations in a lookup table during an offline training phase. This preliminary action eliminates the need for expensive real-time HDI calculations during quote generation, as the system only needs to retrieve pre-computed values based on the customer's attributes and product characteristics.
Solution Approach 2:
The system segments the input variables into discrete categories and creates separate HDI calculations for each combination of segmented variables. By dividing the continuous input space into discrete segments (e.g., customer size categories, product types), the system can pre-compute a manageable number of HDI values that cover all possible scenarios without requiring continuous optimization during runtime.
2Adaptability or versatility
If the system stores and processes a large number of PDF functions for all possible input combinations, then the coverage and applicability of the prediction model are improved, but the storage requirements and data processing complexity increase significantly
Solution Approach 1:
Instead of storing complete PDF function definitions for each input combination, the system creates simplified lookup table entries that copy only the essential HDI results. Each lookup table entry contains pre-computed HDI values that represent the key outcomes of the PDF analysis, eliminating the need to store and process entire probability density functions during runtime.
Solution Approach 2:
The system extracts only the critical HDI information from the complete PDF analysis and stores this extracted data in a compact lookup table format. By taking out only the essential prediction results (the HDI bounds and confidence levels) rather than storing the full PDF mathematical representations, the system achieves both comprehensive coverage and efficient storage.
3Adaptability or versatility
If the system performs real-time HDI calculation for custom input combinations, then the flexibility and customization of quote recommendations are improved, but the response time and computational resources required at runtime increase
Solution Approach 1:
The system performs all complex HDI calculations in advance during the offline training phase, populating a lookup table with pre-computed results for every possible input combination. During real-time quote generation, the system simply retrieves the appropriate pre-computed HDI values from the lookup table based on the customer's attributes, achieving both real-time responsiveness and comprehensive customization without performing expensive calculations at runtime.
Data Source
AI summary
The present disclosure describes a system, method, and computer program for real-time and computationally efficient calculation of a recommended value range for a quote variable, such as price, discount, volume, or closing time. The system uses the highest-density interval (HDI) of probability density function (PDF) as a recommended or suggested value range for a quote variable. PDFs for the quote variable are precomputed for groups of related inputs, and each PDF is summarized as an array of discrete points. A dimension reduction technique is applied to the PDF inputs in both the training and real-time (non-training) phases to reduce the number of possible combinations of PDFs. During a quote-creation process, a PDF look-up table enables the system to efficiently identify an applicable PDF from the group of precomputed PDFs based on reduced input values.