Tensor Parameter Adjustment for Attendance Data Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Graph structure learning techniques, such as deep tensors, often process data with partially similar patterns as common patterns, leading to reduced prediction accuracy, especially in attendance book data where treatment-experienced employees' data can resemble pre-treatment fluctuation patterns, causing noise and inaccuracies in predicting new medical treatments.
Innovation Solution
A machine learning method that generates tensors based on attendance records and modifies parameters for employees on leave and those not on leave, using deep tensors to learn a prediction model by decomposing tensor data and adjusting weights for treatment-experienced employees' data to differentiate it from unwell employees' data, thereby improving prediction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If graph structure learning techniques process attendance data as common patterns, then processing simplicity is maintained, but prediction accuracy deteriorates due to noise from treatment-experienced employees
Solution Approach 1:
The patent applies local quality by differentiating the processing of tensor data based on employee status. Treatment-experienced employees' data is processed with modified parameters (lower weight) compared to other employees' data processed with standard parameters. This allows the system to maintain overall processing simplicity while locally adjusting quality for specific subsets of data that would otherwise introduce noise and reduce prediction accuracy.
2Quantity of substance
If treatment-experienced employees' data is included in training, then data utilization is maximized, but prediction accuracy deteriorates due to pattern similarity with pre-treatment fluctuation
Solution Approach 1:
The patent applies parameter changes by modifying the weight parameter for treatment-experienced employees' data during tensor processing. By changing the weight parameter from the standard value to a modified value (lower weight), the system maintains inclusion of this data in training (maximizing data utilization) while reducing its influence on prediction to prevent accuracy deterioration from pattern similarity with pre-treatment fluctuation.
3Device complexity
If all employee data is processed uniformly, then system complexity is minimized, but prediction accuracy deteriorates due to noise in treatment-experienced employee data
Solution Approach 1:
The patent resolves this contradiction by implementing local quality through status-based differentiation. The system maintains overall simplicity by using a unified tensor processing framework but introduces local quality adjustments specifically for treatment-experienced employees through modified parameters. This allows the system to remain relatively simple while preventing noise from specific data subsets from degrading prediction accuracy.
Data Source
AI summary
A machine learning method includes acquiring data including attendance records of employees and information indicating which employee has taken a leave of absence from work, in response to determining that a first employee of the employees has not taken a leave of absence in accordance with the data, generating a first tensor on a basis of an attendance record of the first employee and parameters associated with elements included in the attendance record, in response to determining that a second employee of the employees has taken a leave of absence in accordance with the data, modifying the parameters, and generating a second tensor on a basis of an attendance record of the second employee and the modified parameters, and generating a model by machine learning based on the first tensor and the second tensor.