Ensemble Learning Prediction Model Feature Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing machine learning ensemble methods fail to accurately differentiate between global and local feature variables, leading to biased reflection of event features in prediction models, which affects prediction accuracy.
Innovation Solution
A computer system is designed to generate prediction models by considering the difference in features indicated by feature variables, using a combination of first and second level training data with meta-features, and applying various machine learning algorithms to enhance prediction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If general machine learning ensemble methods are used, then prediction models can be generated with various algorithms, but the features of events are biased because global and local feature variables are not differentiated
Solution Approach 1:
The patent segments feature variables into two distinct categories: global feature variables (affecting all sample data) and local feature variables (affecting only specific sample data). This segmentation allows the ensemble learning method to treat different feature types appropriately, preventing the bias that occurs when all features are treated uniformly. The segmentation enables separate handling and weighting of global versus local features during model training.
Solution Approach 2:
The patent applies local quality by creating different treatment mechanisms for local feature variables versus global feature variables. Local feature variables are identified as those that have meaningful values only for specific subsets of sample data, and the system adjusts the learning process to give appropriate attention to these localized features rather than applying a uniform approach across all features.
2Reliability
If ensemble learning with multiple prediction models is applied, then prediction accuracy can be improved, but the complexity of the system increases
Solution Approach 1:
The patent segments the ensemble learning process into distinct stages: (1) identification and categorization of global versus local feature variables, (2) generation of multiple prediction models using different algorithms, (3) integration of model predictions with appropriate weighting. This segmentation makes the complex ensemble process more manageable and systematic.
Solution Approach 2:
The patent introduces an intermediary component that identifies and categorizes feature variables as global or local before they are used in model training. This intermediary layer simplifies the overall system by preprocessing feature information and providing structured input to the ensemble learning process, reducing the complexity burden on the model generation and integration stages.
Data Source
AI summary
The prediction accuracy of prediction models generated by ensemble learning is enhanced. A computer system configured to generate a prediction model for predicting an event includes: a storage unit configured to store a plurality of training data including a plurality of sample data including values of a plurality of feature variables and a prediction correct value of the event; and a prediction model generating unit configured to generate a plurality of prediction models using the plurality of training data, to thereby generate a prediction model for calculating an ultimate predicted value on the basis of predicted values of the plurality of prediction models. Prediction models generated by applying the same machine learning algorithm to the plurality of training data are different from each other in features of the event that are reflected in the prediction models.