Forecasting Data Filtering for Limited History and Faster Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional statistical forecast methods require a substantial amount of historical data, typically 3 years, to generate accurate forecasts, which is impractical for rapidly changing environments with limited data availability.
Innovation Solution
An intelligent data filter or algorithm that maximizes the use of available data, allowing accurate forecasting with reduced data storage and increased processing speed by leveraging relevant historical information and index values, especially when full history is lacking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional statistical forecast methods use 3 years of historical data, then forecast accuracy is improved, but data storage capacity requirements increase and processing speed decreases
Solution Approach 1:
The patent extracts only the most relevant historical data points needed for forecasting rather than storing and processing complete multi-year datasets. The system identifies and retains key patterns and trends while discarding redundant information, achieving accurate forecasts with minimal data storage requirements
Solution Approach 2:
The system transforms the forecasting approach by changing parameters from requiring fixed 3-year historical periods to adapting data requirements based on actual data quality, availability, and relevance. This allows the system to achieve accurate forecasts with variable data quantities rather than rigid time-based requirements
2Measurement precision
If traditional statistical forecast methods use 3 years of historical data, then forecast accuracy is improved, but processing speed decreases
Solution Approach 1:
The system extracts only essential data elements required for forecasting calculations, eliminating the need to process entire multi-year datasets. This extraction of critical information maintains forecast accuracy while dramatically reducing computational burden and processing time
Solution Approach 2:
The patent applies partial action by using only the necessary portion of historical data rather than complete multi-year records. The system determines the minimum data quantity and quality needed for accurate forecasts, avoiding excessive data processing while maintaining precision
3Productivity
If data storage capacity is reduced, then processing speed increases, but forecast accuracy may deteriorate
Solution Approach 1:
The system changes the parameter from fixed data quantity requirements to adaptive data selection based on relevance and information content. This allows maintaining forecast accuracy with reduced storage by focusing on high-value data elements rather than volumetric data accumulation
Solution Approach 2:
The patent applies local quality by ensuring high data quality and relevance in the stored subset rather than relying on quantity. The system prioritizes storing high-information-density data points that locally maximize forecasting value while minimizing overall storage requirements
Data Source
AI summary
A method implemented by a computer comprising collecting historical data for a plurality of items; categorizing each of the plurality of items; assigning at least one of the plurality of items as an index item for at least another of the plurality of items; collecting data for a plurality of context parameters related to at least one of the plurality of items; and forecasting a value for one of the plurality of items needed over a future period of time, wherein the method reduces a data storage capacity requirement for the computer and increases said computer's processing speed.