Graph Neural Network Event Prediction Using Sequence Group Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current artificial neural network technologies for predicting subsequent events in sequences fail to consider the temporal order of events, leading to low prediction accuracy.
Innovation Solution
A method and device that generate a sequence group graph based on the time-series order of events, sample this graph to create a computation graph, and use a graph neural network (GNN) to predict subsequent events, incorporating time-series order information to enhance prediction accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If random sampling operation is performed on input data to reduce computational load, then computational efficiency is improved, but temporal order information is lost leading to low prediction accuracy
Solution Approach 1:
The patent segments the event sequence into multiple sub-sequences based on temporal order, creating a structured representation that preserves chronological relationships while enabling efficient processing of smaller sequence segments rather than treating the entire sequence as one unit
Solution Approach 2:
The patent transforms the one-dimensional event sequence into a two-dimensional structure by organizing events into sequences and sub-sequences with hierarchical relationships, adding a structural dimension that preserves temporal order information while facilitating efficient sampling and processing
2Measurement precision
If temporal order of events is considered in sampling operation, then prediction accuracy is improved, but computational load increases
Solution Approach 1:
The patent applies partial sampling to the sequence group graph, selecting only certain nodes and edges based on temporal order criteria rather than processing the entire graph, thus achieving improved prediction accuracy through selective sampling while controlling computational load through the partial action approach
Data Source
AI summary
A method and a device for predicting an event to occur in a device are provided. The method includes identifying a plurality of events that occurred in the device; generating a plurality of sequence graphs based on a time-series order of the plurality of events; generating a sequence group graph by combining the plurality of sequence graphs based on relationships between nodes and edges in the plurality of sequence graphs; obtaining a computation graph by sampling the generated sequence group graph based on time-series order information of the plurality of events; and predicting the event to occur in the device using a graph neural network (GNN) based on the obtained computation graph.