Regression Model Threshold Segmentation for Higher Prediction Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies have not effectively addressed the need for optimizing regression models to reduce residuals and increase predicting accuracy.
Innovation Solution
The proposed solution involves generating a parameter set using Simplification Swarm Optimization rule, which includes a plurality of threshold values and model codes, dividing data into groups, calculating fitness values, and updating best fitness values until a predetermined number of parameter sets is reached.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional regression models are used without optimization, then the modeling process is simple, but the residual errors are large and predicting accuracy is low
Solution Approach 1:
The patent segments the regression modeling process into multiple iterations, where each iteration divides data into different groups using threshold values and applies different regression model codes to different groups. This segmentation allows the system to capture non-linear relationships and reduce residual errors by treating different data segments with appropriate models.
Solution Approach 2:
The patent implements dynamic optimization by iteratively adjusting threshold values and model code assignments based on fitness function evaluations. The Simplification Swarm Optimization dynamically modifies the parameter sets across iterations, transitioning from simple to more sophisticated model configurations as the optimization progresses.
2Measurement precision
If multiple regression models are applied to different data groups, then predicting accuracy improves, but the complexity of model selection and parameter optimization increases
Solution Approach 1:
The patent systematically changes parameters including threshold values, model codes, and data group assignments across multiple iterations. The Simplification Swarm Optimization modifies these parameters based on fitness function feedback, allowing the system to adapt to different data characteristics and achieve better predicting accuracy.
Solution Approach 2:
The patent implements feedback mechanisms through fitness function evaluations that assess the performance of each parameter set. The optimization process uses this feedback to guide subsequent parameter adjustments, creating a closed-loop system that continuously improves model performance based on actual predicting results.
3Manufacturing precision
If iterative optimization with multiple parameter sets is performed, then residual errors are reduced, but the computation time and processing complexity increase
Solution Approach 1:
The patent employs periodic iterative optimization cycles where parameter sets are generated, evaluated, and refined in structured iterations. Each iteration follows a systematic pattern of generating parameter sets, dividing data, calculating fitness values, and updating the best solutions, which balances thorough optimization with computational efficiency.
Solution Approach 2:
The patent performs preliminary actions by pre-defining the optimization framework, including the fitness function, parameter ranges, and iteration structure. This preliminary setup allows the Simplification Swarm Optimization to efficiently navigate the parameter space without requiring extensive trial-and-error adjustments during the optimization process.
Data Source
AI summary
An optimization method of a regression model includes generating a parameter set according to a Simplification Swarm Optimization rule, the parameter set includes a plurality of threshold values and a plurality of model codes, the model codes correspond to a plurality of the regression models, and a plurality of types of the regression models are different from each other; arranging the threshold values; dividing a plurality of data of a dataset into a plurality of groups sequentially according to the threshold values; calculating the data of the groups according to the model codes corresponding to the groups to generate a predicting result and a fitness value of the predicting result; updating a best fitness value in a database according to the fitness value corresponding to the parameter set; and repeating the above steps until a number of the parameter sets being equal to a predetermined value.