Predicting Resources for Unprecedented Workloads via Activity Core Elements
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Predicting resources to fulfill unprecedented workloads is challenging due to uncertainties in capacity, volume, timing, and duration, which can lead to failure, delays, cost escalation, and impacts on other workloads, as existing methods lack accurate forecasting for complex or unestablished workloads.
Innovation Solution
A method involving neural networks that map resource types and capabilities to activity core elements, allowing for the prediction of required resources for unprecedented workloads by decomposing activities into fundamental components and training models to extrapolate necessary resources based on established mappings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing forecasting methods are used for unprecedented workloads, then resource planning can be performed, but prediction accuracy deteriorates due to uncertainties in capacity, volume, timing, and duration
Solution Approach 1:
The patent segments the workload prediction problem into two distinct phases: (1) training phase using historical workload data to establish resource-activity mappings, and (2) prediction phase for unprecedented workloads. This segmentation allows the system to leverage past experience while adapting to new scenarios, resolving the contradiction between prediction accuracy and adaptability to unprecedented workloads.
Solution Approach 2:
The system performs preliminary action by pre-training models on historical workload data before facing unprecedented workloads. The training phase establishes resource capability mappings and activity decompositions in advance, creating a knowledge base that can be rapidly applied to new scenarios without requiring real-time analysis, thus maintaining accuracy while enabling adaptability.
2Adaptability or versatility
If resource capabilities are expanded to handle unprecedented workloads, then workload fulfillment capability improves, but resource costs increase
Solution Approach 1:
The patent changes parameters by decomposing workloads into activity core elements (ACEs) and mapping them to resource capabilities. This transformation allows the system to predict resource needs in terms of specific capabilities rather than raw quantity, enabling optimization of resource selection to match actual workload requirements and avoid unnecessary resource allocation.
Solution Approach 2:
The system replaces traditional mechanical resource allocation methods with machine learning models that predict resource needs based on activity decompositions. This substitution enables more precise resource forecasting, reducing over-provisioning and associated costs while maintaining adequate capability to handle unprecedented workloads.
3Measurement precision
If detailed activity decomposition is performed to improve prediction accuracy, then resource mapping precision improves, but processing complexity increases
Solution Approach 1:
The patent extracts the essential components of workload activities into standardized Activity Core Elements (ACEs). By taking out only the critical activity components needed for resource mapping and eliminating redundant details, the system achieves high prediction precision while reducing the complexity of model training and processing.
Data Source
AI summary
One or more processors receive resource type and capability information and activity information of workloads of a domain. A first model is generated and trained to map the resource information to the activity information of domain workloads. The activity information is decomposed into a set of activity core elements (ACEs). The one or more processors generate a second model, wherein the second model is trained to predict a set of resource types and resource capabilities of the respective resource types, based on an input of the first set of ACEs decomposed from the activity information of the workloads of the domain. The one or more processors receive a second set of ACEs that are decomposed from activities associated with an unprecedented workload, and the one or more processors generate a predicted set of resources to perform the second set of ACEs.