Storage Device Detection Model Training with Dynamic Threshold Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional exceptional storage detection methods are ineffective in various storage scenarios due to differences in workload modes, failing to accurately detect exceptional storage devices.
Innovation Solution
A method is developed to train a device detection model using a test set and training set containing workload data from normal and exceptional storage devices, allowing the model to classify devices based on a threshold degree, with the ability to update this threshold for improved performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional exceptional storage detection methods are used, then the detection process is simple, but the detection accuracy is low due to differences in workload modes across storage scenarios
Solution Approach 1:
The patent applies preliminary action by training a detection model in advance using workload data from multiple storage scenarios before actual detection is needed. The model is pre-trained with diverse workload patterns (sequential, random, mixed) to prepare it for accurate detection across different scenarios without requiring complex real-time adjustments.
Solution Approach 2:
The patent employs parameter changes by adjusting the threshold parameter in the detection model to optimize detection accuracy. The threshold is tuned based on performance metrics (precision, recall, F1-score) to achieve the best balance between detecting exceptional storage devices and minimizing false positives across varying workload modes.
2Reliability
If a fixed threshold degree is used in the detection model, then the model structure is simple, but the performance does not reach threshold performance across different storage scenarios
Solution Approach 1:
The patent applies dynamics by making the threshold parameter adjustable rather than fixed. The threshold can be dynamically tuned based on the specific storage scenario and workload characteristics, allowing the model to adapt to different conditions (sequential, random, mixed workloads) and achieve reliable detection performance across diverse scenarios.
Solution Approach 2:
The patent implements feedback by using detection results and performance metrics to adjust the threshold parameter. The system evaluates detection performance (precision, recall, F1-score) and uses this feedback to optimize the threshold, creating a closed-loop system that continuously improves detection reliability based on actual performance.
3Measurement precision
If the detection model is trained with diverse workload data, then detection accuracy improves, but the data preparation process becomes more complex
Solution Approach 1:
The patent applies universality by creating a detection model that can handle multiple workload modes (sequential, random, mixed) with a single unified training process. The model is designed to be multi-functional, processing diverse workload data types through the same architecture and training methodology, thereby improving detection accuracy without proportionally increasing data preparation complexity.
Data Source
AI summary
Embodiments of the present disclosure relate to a method, an electronic device, and a computer program product for training a model. The method includes: acquiring a test set and a training set for training models, the test set and the training set each including workload data associated with normal storage devices and workload data associated with exceptional storage devices; training a device detection model using the training set, the device detection model being used to classify storage devices as normal storage devices or exceptional storage devices according to a threshold degree, with the threshold degree being within a range; determining a test result by applying the test set to the device detection model; and updating the range of the threshold degree if it is determined that the test result indicates that the performance of the device detection model does not reach a threshold performance. With this method, storage devices can be accurately detected by the trained model.