Dropmessage: Unifying random dropping for graph neural networks
Abstract Graph Neural Networks (GNNs) are powerful tools for graph representation
learning. Despite their rapid development, GNNs also face some challenges, such as over …
learning. Despite their rapid development, GNNs also face some challenges, such as over …
Graph embedding and geometric deep learning relevance to network biology and structural chemistry
P Lecca, M Lecca - Frontiers in Artificial Intelligence, 2023 - frontiersin.org
Graphs are used as a model of complex relationships among data in biological science
since the advent of systems biology in the early 2000. In particular, graph data analysis and …
since the advent of systems biology in the early 2000. In particular, graph data analysis and …
Robust network enhancement from flawed networks
Network data in real-world tends to be error-prone due to incomplete sampling, imperfect
measurements, etc.; this in turn results in inaccurate results when performing network …
measurements, etc.; this in turn results in inaccurate results when performing network …
Minimizing robust density power-based divergences for general parametric density models
A Okuno - Annals of the Institute of Statistical Mathematics, 2024 - Springer
Density power divergence (DPD) is designed to robustly estimate the underlying distribution
of observations, in the presence of outliers. However, DPD involves an integral of the power …
of observations, in the presence of outliers. However, DPD involves an integral of the power …
A stochastic optimization approach to minimize robust density power-based divergences for general parametric density models
A Okuno - arXiv preprint arXiv:2307.05251, 2023 - arxiv.org
Density power divergence (DPD)[Basu et al.(1998), Biometrika], designed to estimate the
underlying distribution of the observations robustly, comprises an integral term of the power …
underlying distribution of the observations robustly, comprises an integral term of the power …
How the latent geometry of a biological network provides information on its dynamics: the case of the gene network of chronic myeloid leukaemia
P Lecca, G Lombardi, RV Latorre… - Frontiers in cell and …, 2023 - frontiersin.org
Background: The concept of the latent geometry of a network that can be represented as a
graph has emerged from the classrooms of mathematicians and theoretical physicists to …
graph has emerged from the classrooms of mathematicians and theoretical physicists to …
Hyperlink regression via Bregman divergence
A Okuno, H Shimodaira - Neural Networks, 2020 - Elsevier
A collection of U (∈ N) data vectors is called a U-tuple, and the association strength among
the vectors of a tuple is termed as the hyperlink weight, that is assumed to be symmetric with …
the vectors of a tuple is termed as the hyperlink weight, that is assumed to be symmetric with …
Graph Embedding with Outlier-Robust Ratio Estimation
K Satta, H Sasaki - IEICE TRANSACTIONS on Information and …, 2022 - search.ieice.org
The purpose of graph embedding is to learn a lower-dimensional embedding function for
graph data. Existing methods usually rely on maximum likelihood estimation (MLE), and …
graph data. Existing methods usually rely on maximum likelihood estimation (MLE), and …
Effective representation learning for graph-structured data with adversarial learning
Q Dai - 2020 - theses.lib.polyu.edu.hk
Graph-structured data is widely existed in real-world applications such as social networks,
paper citation networks and protein-protein interaction networks. It encodes very rich …
paper citation networks and protein-protein interaction networks. It encodes very rich …