Recursive Model Building With Automated Model Sequencing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Enterprise organizations face challenges in implementing machine learning models due to computationally expensive brute force methods and monolithic systems, leading to sub-optimal application across various service lines, with issues such as overfitting, error-proneness, and difficulty in maintaining these systems.
Innovation Solution
A computing platform using machine learning algorithms to automatically select and sequence machine learning models based on problem domains, utilizing microservices architecture for efficient and independent model management, allowing for informed decision-making and reducing computational overhead.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If brute force methods are used to select machine learning models, then model selection is exhaustive, but computational cost increases significantly
Solution Approach 1:
The system uses machine learning algorithms to automatically select and sequence models based on problem domain characteristics, allowing the system to self-optimize without exhaustive brute force computation. The ML algorithm analyzes historical data and service requests to determine optimal model combinations efficiently.
Solution Approach 2:
The system changes the approach from fixed brute force enumeration to dynamic parameter-based selection. It uses problem domain parameters (data volume, complexity, service type) as inputs to an ML algorithm that outputs optimized model sequences, transforming the selection process from computational exhaustion to intelligent inference.
2Device complexity
If monolithic systems are used for model building, then system integration is simple, but maintenance difficulty and cost increase
Solution Approach 1:
The patent implements a microservices architecture where the model building system is divided into independent, loosely-coupled services. Each microservice handles specific model building tasks and can be maintained, updated, or scaled independently without affecting the entire system, thus reducing maintenance difficulty while preserving integration through standardized communication protocols.
3Device complexity
If monolithic systems are used, then system structure is simple, but system downtime during maintenance increases
Solution Approach 1:
By segmenting the system into microservices, the patent enables independent maintenance of individual services without shutting down the entire system. Other microservices continue to operate during maintenance windows, significantly reducing overall system downtime while maintaining a manageable structural complexity through standardized service communication patterns.
4Adaptability or versatility
If brute force methods are applied to all service lines, then model coverage is comprehensive, but processing resources are wasted
Solution Approach 1:
The system applies different model selection strategies to different service lines based on their specific characteristics. The ML algorithm analyzes each service's data volume, complexity, and requirements to determine the optimal model sequence, avoiding unnecessary brute force computation on simple services while maintaining comprehensive coverage through targeted, quality-adjusted model application.
Data Source
AI summary
Aspects of the disclosure relate to an automated model generation platform for recursive model building. A computing platform may receive a request for automated machine learning model building, and may identify a service offering corresponding to the request. Based on the identified service offering and using machine learning algorithms, the computing platform may select machine learning models and a corresponding sequence of model application (e.g., machine learning model information). The computing platform may store the machine learning model information along with a corresponding indication of the identified service offering. The computing platform may receive a request for model information corresponding to a service access request, and may identify that the service access request corresponds to a problem within the identified service offering. In response, the computing platform may send the machine learning model information, which may cause the enterprise service host system to generate a service output interface.