ML Model Aggregation with Multiplier for Prediction Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Machine learning models require extensive processing time for training and retraining, especially when data associated with specific variables changes, leading to inefficient use of computational resources and outdated predictions due to infrequent updates.
Innovation Solution
Training multiple distinct machine learning models using split datasets with identical or non-identical features, and aggregating their predictions with a machine learning multiplier model that applies industry-specific indicators to generate a final prediction, allowing for faster updates and more recent data incorporation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single machine learning model is trained using all available data, then the model can provide comprehensive predictions, but the training requires extensive processing time and computational resources
Solution Approach 1:
The patent divides the single comprehensive model into multiple specialized models, each trained on a specific subset of data (e.g., different time periods, data sources, or feature sets). This segmentation reduces the training time and computational resources required for each individual model while maintaining comprehensive prediction coverage through aggregation of multiple model outputs.
2Measurement precision
If machine learning models are retrained frequently to incorporate recent data trends, then prediction accuracy improves, but computational costs and processing time increase significantly
Solution Approach 1:
By segmenting the model training into multiple smaller models trained on different data subsets, the system can update individual models more efficiently and frequently without the prohibitive computational cost of retraining a single large comprehensive model. This enables more frequent incorporation of recent data trends at reduced computational expense.
Solution Approach 2:
The system implements periodic updates of individual model components rather than continuous retraining of the entire system. This allows for scheduled incorporation of recent data trends into specific models based on their data subsets, balancing prediction accuracy with computational resource management.
3Adaptability or versatility
If multiple machine learning models are trained and their predictions aggregated, then the system can incorporate diverse data sources and update more frequently, but the system complexity increases
Solution Approach 1:
The system segments the prediction task into multiple specialized models, each handling specific data sources or feature sets. This segmentation naturally increases adaptability to diverse data sources while organizing complexity into manageable, modular components that can be independently trained and updated.
Solution Approach 2:
The patent combines the outputs of multiple specialized models through aggregation to produce a unified prediction. This merging approach maintains the benefits of diverse data sources and frequent updates while presenting a simplified interface to users, effectively managing system complexity through structured integration.
Data Source
AI summary
Systems, methods, and computer-readable storage media for aggregating the outputs of multiple machine learning models, then using the output of yet another machine learning model as a multiplier to obtain a final prediction. A system can receiving a plurality of data sets, each data set being associated with at least one data type, and train machine learning models, each model associated with one or more of the different data types. Upon execution, the multiple machine learning models can each produce a prediction which is aggregated together to form an aggregated prediction. The multiplier from the additional machine learning model can then be applied to the aggregated prediction, resulting in a final prediction.