Edge-Cloud Machine Learning Data Selection via Segmented Evaluation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge is to efficiently select effective input data for relearning in a machine learning system where resource-limited edge devices struggle with large arithmetic operations, leading to increased communication burdens when transmitting data to the cloud for processing.
Innovation Solution
A machine learning system with edge and cloud devices that calculate and compare evaluation values for input data using different standards to select candidate data for relearning, with the edge device calculating a first evaluation value based on data correlation and the cloud device calculating a second evaluation value based on inference results, allowing for efficient data selection and reduction of unnecessary data transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the edge device executes selection processing requiring large arithmetic amount, then the effectiveness of input data for relearning can be improved, but the resource limitations of the edge device make it difficult to perform such processing
Solution Approach 1:
The patent divides the selection processing into two segments: edge device executes first selection based on simple correlation calculation, cloud device executes second selection based on comprehensive effectiveness evaluation. This segmentation allows complex processing to be performed on cloud while edge device handles only lightweight preliminary filtering.
2Measurement precision
If a large amount of input data is transmitted to the cloud, then the cloud can execute selection processing requiring large arithmetic amount, but the communication burden on the machine learning system is greatly increased
Solution Approach 1:
The patent extracts and transmits only the essential correlation information from the edge device to the cloud, rather than transmitting all raw input data. This extraction approach enables the cloud to perform accurate effectiveness evaluation while minimizing data transmission volume.
3Quantity of substance
If the edge device performs comprehensive selection processing, then the communication burden is reduced, but the resource limitations prevent the edge device from executing such processing
Solution Approach 1:
The patent segments the selection processing into two stages performed by different devices. The edge device performs simple first selection with low computational requirements, filtering out obviously ineffective data. The cloud device performs comprehensive second selection on the reduced dataset. This segmentation reduces both transmission volume and edge device complexity requirements.
Data Source
AI summary
A machine learning system includes first and second information processing devices. The first-information-processing device includes a first-evaluation unit, a first-selection unit, and a candidate-data-transmission unit. The first-evaluation unit calculates a first evaluation value for each of candidate data pieces based on a first evaluation standard. The first-selection unit selects whether each input data is included in the candidate data pieces based on the first evaluation value. The candidate-data-transmission unit transmits the candidate data. The second-information-processing device includes a candidate-data-reception unit, a second-evaluation unit, and a second-selection unit. The candidate-data-reception unit receives the candidate data. The second-evaluation unit calculates a second evaluation value for each candidate data based on a second evaluation standard different from the first evaluation standard. The second-selection unit selects whether each candidate data is included in learning data pieces based on the second evaluation value.