Virtual node graph neural network for full phonon prediction.
Ryotaro OkabeAbhijatmedhi ChotrattanapitukArtittaya BoonkirdNina AndrejevicXiang FuTommi S. JaakkolaQichen SongThanh NguyenNathan C. DruckerSai MuYao WangBolin LiaoYongqiang ChengMingda LiPublished in: Nat. Comput. Sci. (2024)
Keyphrases
- neural network
- graph structure
- prediction model
- directed graph
- prediction accuracy
- artificial neural networks
- graph representation
- undirected graph
- multi layer perceptron
- elman network
- graph model
- connected components
- edge weights
- root node
- virtual environment
- prediction algorithm
- overlapping communities
- nodes of a graph
- weighted graph
- genetic algorithm
- random walk
- fuzzy logic
- bp neural network
- directed acyclic graph
- neural network is trained
- betweenness centrality
- spanning tree
- virtual world
- path length
- graph theory
- tree structure
- structured data
- finding the shortest path
- graph databases
- network architecture
- prediction error
- activation function
- graph matching
- protein interaction
- back propagation
- strongly connected
- multilayer perceptron
- chaotic time series
- recurrent neural networks
- graphical models
- bipartite graph