ML Application Recommendation Using Task-Based Performance Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users are burdened by the overwhelming number of machine learning (ML) service options, making it difficult to choose the best service for each input instance of a target task, leading to inefficiencies in task execution.
Innovation Solution
A system and method that utilize machine learning to analyze task data, determine performance scores for various software applications, and recommend the most suitable application based on these scores, thereby streamlining the task execution process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If users have access to multiple machine learning service options, then the quality and variety of available solutions improve, but the complexity of choosing the best service increases
Solution Approach 1:
The patent introduces an intermediary system that automatically evaluates multiple ML services and recommends the best one for each input instance. This mediator analyzes service outputs, compares performance, and guides user selection, thereby resolving the contradiction by maintaining service variety while eliminating selection complexity through automated assistance.
2Measurement precision
If users manually evaluate each ML service for every task, then the accuracy of service selection improves, but the time and effort required increases
Solution Approach 1:
The system performs preliminary evaluation of multiple ML services in advance by processing sample inputs through each service and comparing outputs before the user needs to make a selection. This pre-computed performance data is stored and reused for future similar tasks, achieving high selection accuracy without requiring users to spend time on repeated manual evaluations.
Solution Approach 2:
The system enables self-service by automatically performing the service evaluation and selection process without user intervention. The automated system independently compares ML service outputs, determines performance metrics, and recommends the best service, thereby achieving accurate selection while eliminating the time users would otherwise spend on manual evaluation.
3Ease of operation
If a single ML service is used for all tasks, then the simplicity of the system improves, but the ability to optimize for each specific task decreases
Solution Approach 1:
The patent implements a dynamic service selection mechanism that automatically adapts the ML service choice based on the specific characteristics of each input instance and task requirements. The system evaluates multiple services and dynamically selects the most appropriate one for each case, thereby maintaining simple user interaction while achieving optimized task execution efficiency through adaptive service selection.
Data Source
AI summary
In some examples, a server instructs individual software applications to process individual tasks and determines a plurality of outputs resulting from processing. The server determines, based on the plurality of outputs, individual performance scores associated with individual software applications, and determines individual features associated with individual task data of multiple task data. The server receives task data associated with a task, determines at least one feature associated with the task data based on analyzing the task data and predicts, using at least one machine learning model, the individual performance scores associated with the individual software applications that have processed the task. The server selects at least one software application from the plurality of software applications based on an associated performance score, generates a recommendation of the at least one software application, and transmits the recommendation to the user device.