Static-Dynamic AI Layers for Progressive Data Pipeline Decisions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional AI systems make static decisions based on data at a certain point in time, leading to unnecessary computing resource usage and inefficiency due to the inability to generate dynamic decisions for changing data.
Innovation Solution
A multi-layer AI framework that includes separate models for static and dynamic features, allowing for progressive decision-making by processing static features once and updating decisions based on changing dynamic features over time, reducing computational load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional AI systems process all features (static and dynamic) at every decision point, then comprehensive decisions can be made, but computing resource usage increases unnecessarily
Solution Approach 1:
The patent segments the feature processing into two distinct layers: a first layer that processes static features once and stores the results, and a second layer that processes only dynamic features at subsequent decision points. This segmentation eliminates redundant processing of static features while maintaining comprehensive decision-making capability by combining results from both layers.
2Loss of time
If AI systems reprocess all features at each time point, then up-to-date decisions are generated, but processing time and computational load increase
Solution Approach 1:
The patent applies preliminary action by processing static features in advance (at the first time point) and storing the results for later reuse. When dynamic features change at subsequent time points, the system only needs to process the changed dynamic features and combine them with the pre-processed static feature results, significantly reducing processing time and computational load while maintaining up-to-date decision accuracy.
Data Source
AI summary
There are provided systems and methods for multi-layer artificial intelligence models for progressive predictions during data pipeline management. A service provider may provide AI functionalities, such as through a multi-layer ML model framework that employs multiple layers for different ML models that process different features. The features in one layer and ML model may process data for static features, where an output from this layer may be used as an input with data for dynamic feature that provide a predictive score or output for the input data. The static features may only be required to be processed once or a few times in the first layer and may not be required to be further processed again at later times. With the second layer, the data for the dynamic features may change, and thus the second layer may process new data without being required to reprocess the static features.