AutoML Framework for Offline-Online Self-Learning Deployment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing AutoML frameworks are limited to static, non-customizable predictive components and lack comprehensive data science processes that integrate offline and online data analysis, require significant manual work, and do not support various machine learning libraries.
Innovation Solution
A unified, customizable, and extensible framework for automated data science processes that include descriptive, exploratory, predictive, prescriptive, automation, and autonomous components, integrating offline and online data analysis, and supporting multiple machine learning libraries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If conventional data science processes are used, then manual work can be performed with existing tools, but the process is not comprehensive enough to support value-driven tasks and requires significant human intervention
Solution Approach 1:
The patent implements a unified AutoML framework that integrates multiple data science components (descriptive, exploratory, predictive, and prescriptive analytics) into a single system. This framework can automatically perform various data science tasks across different domains and applications, making it universally applicable while maintaining comprehensiveness through its modular architecture that supports multiple machine learning libraries and methodologies
2Ease of manufacture
If static AutoML frameworks are used, then predictive modeling can be automated for specific implementations, but the frameworks are not generally customizable or extensible
Solution Approach 1:
The patent implements a dynamic and flexible AutoML framework that allows users to customize and extend the system according to specific needs. The framework supports dynamic configuration of machine learning pipelines, enables integration of different machine learning libraries, and provides extensible interfaces for incorporating domain-specific requirements, making it adaptable to various applications while maintaining ease of deployment through automated processes
3Productivity
If only predictive components are handled, then focused automation can be achieved, but other components of the data science process remain manual
Solution Approach 1:
The patent merges multiple previously separate data science components (descriptive analytics, exploratory analytics, predictive analytics, and prescriptive analytics) into a unified automated framework. This integration allows the system to automatically perform end-to-end data science workflows, combining various analytical functions into a cohesive process that improves overall productivity while extending automation beyond just predictive modeling to include all stages of the data science lifecycle
4Measurement precision
If offline processes are used, then comprehensive analysis can be performed on historical data, but online real-time data science processes are not supported
Solution Approach 1:
The patent segments the data science process into distinct offline and online components that can operate independently and in coordination. The offline process handles comprehensive analysis of historical data with full computational resources, while the online process enables real-time data science operations with optimized performance. This segmentation allows the system to maintain high accuracy through thorough offline analysis while achieving real-time processing capabilities through streamlined online operations
Data Source
AI summary
Example implementations described herein are directed to systems and methods for generation and deployment of automated and autonomous self-learning machine learning models, which can include generating a predictive model and a prescriptive model through an offline learning process at a first system; controlling operations of a second system through deploying the predictive model and the prescriptive model to the second system; and autonomously updating the predictive model and the prescriptive model from feedback from the second system through an online learning process while the prescriptive model and the predictive model are deployed on the second system.