Client-Side ML Prediction for Selective Data Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Optimizing the use of various types of memory storage in computing, especially in cloud computing, is challenging due to a lack of understanding about data retrieval and update frequencies, and limited access to certain types of information that are restricted to local client devices.
Innovation Solution
Using a machine learning model to predict future events based on update data, allowing client devices to select an appropriate data store by generating time-related predictions that satisfy specific criteria associated with different data stores.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If server-side machine learning models are used for data storage optimization, then centralized control is improved, but computational requirements and energy consumption increase
Solution Approach 1:
The system divides the machine learning model into two segments: a training model that runs on the server to generate predictions, and a client-side model that receives these predictions and makes local decisions about data storage. This segmentation allows the server to perform only the computationally intensive training operation periodically, while the client device handles real-time prediction and decision-making, significantly reducing server computational requirements and energy consumption while maintaining centralized control over the overall system behavior.
2Measurement precision
If more data is collected for prediction accuracy, then prediction accuracy is improved, but data privacy and security risks increase
Solution Approach 1:
The server acts as an intermediary that receives anonymized or aggregated data from client devices, processes it through the training model to generate predictions, and sends only the prediction results back to the client without exposing raw data. This intermediary mechanism allows the system to improve prediction accuracy by collecting data from multiple sources while maintaining data privacy and security, as the actual sensitive data never leaves the client device and the server only processes aggregated information.
3Productivity
If multiple data stores are used for optimization, then storage efficiency is improved, but system complexity increases
Solution Approach 1:
The client device is equipped with a local model that autonomously evaluates multiple data store options and automatically selects the most appropriate one based on real-time conditions and server-generated predictions. This self-service capability allows the system to achieve high storage efficiency by dynamically choosing optimal data stores without requiring complex manual configuration or centralized decision-making for each storage operation, thereby reducing system complexity while maintaining optimization benefits.
Data Source
AI summary
A method and related system for storing data based on predictions includes obtaining, from a client device, a first set of update data and a time-related prediction generated by a client-side version of a machine learning model. The method further includes determining that the time-related prediction is associated with a first data store of a plurality of data stores comprising the first data store and a second data store. The method further includes updating a record in the first data store based on the first set of update data in response to a determination that the time-related prediction is associated with the first data store. The method further includes updating the record in the first data store based on a second set of update data obtained after obtaining the first set of update data.