Reverse Forecasting for Time Series Data Extension
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional forecasting algorithms, such as the Holt-Winters algorithm, require a sufficient length of time series data to function effectively, which is often not available due to incomplete historical data, leading to the inability to perform forecasting in various business and organizational contexts.
Innovation Solution
A technique that estimates additional data points by reversing the time order of initial time series data and applying a forecasting algorithm to produce past forecast data, which is then combined with the initial data to extend the series to a sufficient length, allowing the application of forecasting algorithms like Holt-Winters or seasonal ARIMA.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional forecasting algorithms are applied, then forecasting accuracy is improved, but data length requirement causes inability to perform forecasting when historical data is insufficient
Solution Approach 1:
The patent applies reverse forecasting by inverting the time order of existing data points. Instead of forecasting forward from insufficient historical data, the algorithm forecasts backward to generate synthetic historical data points. This inversion allows the system to create the required data length by generating pseudo-historical values that maintain the statistical properties and patterns of the original data, enabling conventional forecasting algorithms to be applied successfully.
2Productivity
If historical data collection is delayed or incomplete, then data availability is reduced, but forecasting capability is lost
Solution Approach 1:
The patent performs preliminary action by generating synthetic historical data points before the actual forecasting process. By creating these pseudo-historical values in advance through reverse forecasting, the system prepares sufficient data length upfront, eliminating the need to wait for complete historical data collection. This preliminary data generation enables immediate forecasting capability even when historical data collection is delayed or incomplete.
Data Source
AI summary
To forecast data, an initial collection of data having a first length is received. In response to determining that the first length of the initial collection of data is insufficient for performing forecasting using a forecasting algorithm, an order of the initial collection of data is reversed to provide a reversed collection of data. Forecasting is applied on the reversed collection of data to estimate additional data values to combine with the initial collection of data to provide a second collection of data having a second length greater than the first length. The forecasting algorithm is applied on the second collection of data.