Multi-Granularity Data Forecasting for Outsourcing Planning

Solution Overview

Problem

Existing outsourcing service planning systems face inconsistencies in regional outsourcing service plans when forecasting and planning are performed in different dimensions, leading to uncertainty in decision-making.

Innovation Solution

A data processing method that generates and models data sets of different granularities, using historical data to create a second-granularity forecasting model, which is then corrected and smoothed to ensure consistency with first-granularity data, and further refined through feedback correction, ensuring precise forecasting.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If forecasting and planning are performed directly in a region dimension to obtain a regional outsourcing service plan, then the forecasting process is simplified, but the forecast data may lack detailed project-level accuracy

Engineering Contradiction:

Improveforecasting process complexityVSAvoidforecast data accuracy

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The patent segments the forecasting process into two distinct granularities: region-level forecasting and project-level forecasting. By dividing the overall forecasting task into these segments, the system can generate both high-level regional plans and detailed project-specific plans, resolving the contradiction between process simplicity and data accuracy.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces a granularity dimension to the forecasting system, allowing forecasts to be generated at multiple levels (region level and project level). This dimensional approach enables the system to maintain both simplified regional forecasting and detailed project forecasting simultaneously, addressing the contradiction between complexity and accuracy.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

2Measurement precision

If forecasting and planning are performed in a project dimension to obtain outsourcing service plans of all projects in a region, then detailed project-level forecast accuracy is improved, but the forecasting process complexity increases

Engineering Contradiction:

Improveforecast data accuracyVSAvoidforecasting process complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the forecasting process into two distinct granularities: region-level forecasting and project-level forecasting. By dividing the overall forecasting task into these segments, the system can generate both high-level regional plans and detailed project-specific plans, resolving the contradiction between process simplicity and data accuracy.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces a granularity dimension to the forecasting system, allowing forecasts to be generated at multiple levels (region level and project level). This dimensional approach enables the system to maintain both simplified regional forecasting and detailed project forecasting simultaneously, addressing the contradiction between complexity and accuracy.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

3Quantity of substance

If regional outsourcing service plans from different forecasting dimensions are aggregated, then comprehensive coverage is achieved, but consistency between plans of different granularities deteriorates

Engineering Contradiction:

Improveforecast data coverageVSAvoiddata consistency

Core Design Contradiction:

Quantity of substanceVSStability of the object's composition

Solution Approach 1:

The patent implements a feedback mechanism where project-level forecast results are aggregated to region-level and compared with direct region-level forecasts. The system uses this feedback to identify and resolve inconsistencies, ensuring that both granularities produce consistent results. This feedback loop maintains data consistency while achieving comprehensive coverage.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent merges the region-level forecast and project-level forecast into a unified forecasting model. By combining both forecasting approaches and ensuring they produce consistent results, the system achieves comprehensive coverage while maintaining data consistency across different granularities.

Inventive Principle:

Principle #5Merging (Combining)

Data Source

PatentUS10169290B2Data processing method and apparatus

Publication Date: 2019.01.01 HUAWEI TECH CO LTD

AI summary

The present invention provides a data processing method, including: acquiring historical data, where the historical data belongs to a first level and a second level, and data corresponding to the first level comprises data corresponding to the second level; generating, from the historical data, a first-granularity data set according to a first granularity, and generating, from the historical data, a second-granularity data set according to a second granularity, where the first granularity and the second granularity respectively correspond to the first level and the second level; performing modeling for a second-granularity forecasting model according to the first-granularity data set and the second-granularity data set; and performing forecasting by using the second-granularity forecasting model to obtain second-granularity forecast data. The present invention enables obtained forecast data of different granularities to be consistent.