Embedding Map for Resource Availability Notifications
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional systems have not effectively leveraged explainability vectors from machine learning models to identify similar user systems for resource availability notifications, lacking a method to generate embedding maps that translate user profiles into a meaningful embedding space.
Innovation Solution
The method involves extracting explainability vectors from a first machine learning model, using these vectors to generate embedding maps that translate user profiles into an embedding space, and then processing these embeddings with a second machine learning model to cluster and identify similar user systems for resource availability notifications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional systems use machine learning models for predicting resource consumption, then prediction capability is achieved, but the ability to identify similar user systems for notifications is lost
Solution Approach 1:
The patent introduces an embedding map as an intermediary that translates user profiles into an embedding space. This embedding space serves as a mediator between the machine learning model's prediction capability and the system's ability to identify similar user systems. The embedding map converts raw user profile data into a transformed space where similarity can be effectively measured and clusters can be formed, thus enabling both prediction and identification of similar systems simultaneously.
2Loss of information
If explainability vectors are extracted from machine learning models, then interpretability is improved, but direct use for resource allocation adjustment is limited
Solution Approach 1:
The patent transforms the explainability vectors by applying an embedding map that changes the parameter space. Instead of using the raw explainability vectors directly for resource allocation, the system transforms them into an embedding space where they can be effectively utilized for identifying similar user systems and forming clusters. This parameter transformation enables the explainability information to be practically applied in a different but more useful context.
3Measurement precision
If embedding maps are generated using explainability vectors, then accuracy in identifying similar user systems is improved, but system complexity increases
Solution Approach 1:
The patent performs preliminary action by pre-computing the embedding map from the explainability vectors before the actual similarity identification process. This embedding map is generated in advance and can be reused for multiple similarity queries, thus reducing the computational complexity during runtime. The heavy lifting of creating the embedding space is done once beforehand, making subsequent operations simpler and faster.
Data Source
AI summary
Systems and methods for executing resource availability notifications to user systems are described. In some aspects, the system receives, for a first plurality of user systems, a first plurality of user profiles and a plurality of resource availability values. Each user profile includes values for a set of features. The system processes a first machine learning model which generates resource availability values from the set of features and extracts an explainability vector. The system uses the explainability vector to generate an embedding map that translates feature values into a corresponding embedding in an embedding space. The system encodes a second plurality of user profiles and processes the resulting user profile vectors using a second machine learning model to generate clusters of user profile vectors. The system selects a cluster from the clusters of user profile vectors and determines user systems corresponding to the cluster for executing resource availability notifications.