Hierarchical Modeling Node for Visual Forecasting Pipelines
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing visual forecasting software lacks the ability to efficiently create custom time-series hierarchies and corresponding level pipelines within a graphical user interface, limiting user customization and flexibility in forecasting processes.
Innovation Solution
The introduction of a hierarchical modeling node in the graphical user interface allows users to define custom time-series hierarchies with multiple levels, enabling separate customization of level pipelines for each hierarchy level independently. This node facilitates the creation of a default level pipeline for each hierarchy level, which can be automatically updated based on changes to the hierarchy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If existing visual forecasting software is used, then the forecasting process can be executed, but users cannot efficiently create custom time-series hierarchies and corresponding level pipelines within the graphical user interface
Solution Approach 1:
The forecasting system is segmented into multiple hierarchical levels, where each level represents a distinct time-series hierarchy. Users can create and customize separate level pipelines for each hierarchy level independently, allowing granular control over forecasting processes at different aggregation levels without affecting other levels.
Solution Approach 2:
The system introduces a hierarchical dimension to the traditional flat forecasting pipeline. By organizing forecasting operations across multiple levels (e.g., daily, weekly, monthly aggregations), users can navigate and customize pipelines in a multi-dimensional space, enabling complex forecasting scenarios while maintaining interface organization.
2Adaptability or versatility
If users want to customize separate level pipelines for each hierarchy level, then flexibility increases, but the complexity of managing multiple pipelines increases
Solution Approach 1:
The graphical user interface implements universal pipeline management capabilities that work across all hierarchy levels. The same interface mechanisms (drag-and-drop, configuration panels, execution controls) are reused for creating, modifying, and running pipelines at any level, reducing the learning curve and operational complexity despite the increased number of pipelines.
Solution Approach 2:
Level pipelines are nested within the hierarchical structure, where each level's pipeline is contained within the broader hierarchical context. This nesting allows users to manage pipelines in a hierarchical manner, where higher-level pipelines can reference or aggregate results from lower-level pipelines, organizing complexity in a structured way.
3Adaptability or versatility
If a hierarchical modeling node is added to enable custom hierarchies, then user flexibility improves, but the device complexity increases
Solution Approach 1:
The hierarchical modeling node acts as an intermediary component in the graphical user interface that mediates between the user's hierarchy definition needs and the underlying forecasting engine. This node provides a standardized interface for defining time-series hierarchies, translating user-friendly hierarchy specifications into the internal representation required for executing level pipelines.
Data Source
AI summary
A system described herein can generate a graphical user interface (GUI) for a piece of forecasting software. The GUI can include graphical nodes arranged on a drag-and-drop canvas to define an overall forecasting pipeline. The graphical nodes can include a hierarchical modeling node that enables a user to define a time series hierarchy. The hierarchical modeling node can also enable separate level pipelines to be customized for each level of the time series hierarchy. The level pipelines can form subparts of the overall forecasting pipeline. The system can then execute the level pipelines to generate multiple forecasts, where each forecast corresponds to a respective level of the time series hierarchy. In some examples, the system can execute a reconciliation process on the forecasts to generate reconciled forecasts. Each reconciled forecast can correspond to one of the forecasts. The system may then generate one or more visualizations of the reconciled forecasts.