Personalized Automated Machine Learning Model Recommendation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automated machine learning systems currently available in the marketplace are unable to generate machine learning models that accommodate individual user preferences or domain-specific requirements, as they do not consider the varying importance of raw data across different industries, leading to generic model recommendations for all users.
Innovation Solution
A method and apparatus utilizing three neural networks to personalize machine learning models by classifying users and models, and predicting relevance scores for each user-model combination, allowing for tailored model recommendations based on user profiles, model profiles, and user-model interaction data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If automated machine learning systems generate the same list of machine learning models for all users, then the system complexity is reduced and processing time is decreased, but the models fail to accommodate individual user preferences and domain-specific requirements
Solution Approach 1:
The system segments users into different groups based on their preferences and domain characteristics using clustering algorithms. This segmentation enables the system to provide personalized model recommendations for each user group while maintaining manageable system complexity through grouped processing rather than individual customization for each user.
Solution Approach 2:
The system performs preliminary analysis of user preferences and domain characteristics before generating model recommendations. By pre-processing user data and identifying key preference patterns in advance, the system prepares personalized recommendations that adapt to individual users without adding complexity to the core model generation process.
2Measurement precision
If automated machine learning systems consider individual user preferences and domain-specific requirements, then model relevance and accuracy are improved, but processing time and computational resources increase
Solution Approach 1:
The system implements a two-stage recommendation process where it first provides a set of candidate models and then refines recommendations based on user feedback and interaction data. This partial action approach allows the system to deliver acceptable recommendations quickly while having the capability to improve accuracy over time, balancing processing time with model recommendation accuracy.
Solution Approach 2:
The system incorporates feedback mechanisms that track user interactions with recommended models and use this information to refine future recommendations. By learning from user feedback, the system improves model recommendation accuracy for individual users while reducing the computational burden of analyzing all possible models from scratch each time.
3Adaptability or versatility
If automated machine learning systems analyze user preferences and domain characteristics to personalize recommendations, then user satisfaction is improved, but the data processing requirements and system resources increase
Solution Approach 1:
The system extracts and focuses on the most relevant features from user preference data and domain characteristics rather than processing all available data. By identifying and extracting key discriminatory features that most strongly correlate with model preferences, the system achieves effective personalization while significantly reducing data processing requirements.
Solution Approach 2:
The system applies different levels of data analysis and processing to different user segments and domains based on their specific characteristics. Rather than uniformly processing all user data at the same level, the system tailors the depth and type of data analysis to local requirements, optimizing resource usage while maintaining personalization capability.
Data Source
AI summary
In accordance with an embodiment of the invention, a method is provided for personalizing machine learning models for users of an automated machine learning system, the machine learning models being generated by an automated machine learning system. The method includes obtaining a first set of datasets for training first, second, and third neural networks, inputting the training datasets to the neural networks, tuning hyperparameters for the first, second, and third neural networks for testing and training the neural networks, inputting a second set of datasets to the trained neural networks and the third neural network generating a third output data including a relevance score for each of the users for each of the machine learning models, and displaying a list of machine learning models associated with each of the users, with each of the machine learning models showing the relevance score.