Feature Graph Engine for ML Model Feature Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current feature selection methods for machine learning models in network systems are time-consuming, complex, and often require manual expertise or statistical approaches, which can lead to inefficiencies and increased resource usage.
Innovation Solution
The use of Feature Graph Engines (FGEs) to derive a design-time set of features for network ML models, minimizing redundant data and optimizing resource usage by creating feature graphs that represent network layers and update dynamically based on runtime changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional statistical approaches are used for feature selection, then feature selection can be performed, but the process is time-consuming and complex
Solution Approach 1:
The patent replaces traditional statistical mechanical approaches with a graph-based system. Feature graphs model relationships between features using graph theory, and graph algorithms automatically identify important features through structural analysis rather than complex statistical computations, significantly reducing processing time and complexity
Solution Approach 2:
The patent segments the feature selection process into distinct graph construction phases and feature identification phases. By dividing the complex task into manageable steps (building feature graphs, analyzing graph structures, selecting important features), the system achieves better computational efficiency and scalability
2Measurement precision
If more features are included in ML models, then prediction accuracy may improve, but computational cost and power consumption increase
Solution Approach 1:
The patent extracts and selects only the most important features from the complete feature set using graph-based importance analysis. By identifying and extracting only the essential features that contribute most to prediction accuracy, the system maintains high prediction quality while significantly reducing the number of features processed, thereby lowering computational cost and energy consumption
Solution Approach 2:
The patent changes the parameter representation from raw feature values to graph-structured representations. By transforming features into graph models where relationships and importances are encoded structurally, the system enables more efficient processing and selection, reducing computational requirements while maintaining or improving prediction accuracy
3Reliability
If manual expertise is used for feature selection, then feature selection can be optimized, but the process becomes more complex and time-consuming
Solution Approach 1:
The patent implements self-service feature selection where the system automatically builds feature graphs and identifies important features without requiring manual expertise. The graph-based algorithms autonomously analyze data relationships, construct feature graphs, and select optimal features, eliminating the need for manual intervention while maintaining high selection quality
Solution Approach 2:
The patent substitutes manual expert analysis with automated graph-based computational systems. By using graph theory and algorithmic analysis to replace manual feature selection processes, the system achieves consistent, reproducible results without the time consumption and subjectivity inherent in manual approaches
Data Source
AI summary
The present disclosure relates to systems and methods for ML model feature selection and transformation. Specifically, the system and method Include receiving information and data from a network having resources; implementing feature selection on one or more network Machine Learning (ML) models, such that each is a pipeline of a plurality of functions to control the resources and with specified interfaces to other control applications; utilizing one or more feature graph engines (FGEs) which creates one or more feature graphs, from the information and data, as a functional component to derive a design-time set of feature vector for a specific context, each feature graph represents network layer representations in the network which includes multiple layers; and implementing changes to the one or more feature graphs based on any run-time updates from the pipeline.