Federated learning of molecular properties with graph neural networks in a heterogeneous setting

Zhu, Wei; Luo, Jiebo; White, Andrew D.

doi:10.1016/j.patter.2022.100521

Cited by 17 publications

(4 citation statements)

References 48 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…One potential solution is Federated Learning. Federated Learning is an approach that allows for model training across multiple decentralized devices or servers holding local data samples, without data exchange [101,102]. This not only preserves data privacy but also facilitates the collaborative nature of AI without centralizing sensitive information.…”

Section: Discussionmentioning

confidence: 99%

Transforming organic chemistry research paradigms: Moving from manual efforts to the intersection of automation and artificial intelligence

Liu,

Chen,

2023

NSO

View full text Add to dashboard Cite

Organic chemistry is undergoing a major paradigm shift, moving from a labor-intensive approach to a new era dominated by automation and artificial intelligence (AI). This transformative shift is being driven by technological advances, the ever-increasing demand for greater research efficiency and accuracy, and the burgeoning growth of interdisciplinary research. AI models, supported by computational power and algorithms, are drastically reshaping synthetic planning and introducing groundbreaking ways to tackle complex molecular synthesis. In addition, autonomous robotic systems are rapidly accelerating the pace of discovery by performing tedious tasks with unprecedented speed and precision. This article examines the multiple opportunities and challenges presented by this paradigm shift and explores its far-reaching implications. It provides valuable insights into the future trajectory of organic chemistry research, which is increasingly defined by the synergistic interaction of automation and AI.

show abstract

Section: Discussionmentioning

confidence: 99%

Transforming organic chemistry research paradigms: Moving from manual efforts to the intersection of automation and artificial intelligence

Liu,

Chen,

2023

NSO

View full text Add to dashboard Cite

show abstract

“…So far, existing FGL studies can be categorized into three types, i.e., inter-graph, intra-graph, and graph-structured FGL (Zhang et al 2021a). In intergraph FGL, each client owns a set of graphs and participates in FL training to learn a better GNN to model local data (Xie et al 2021), learn a generalizable model (Zhu, Luo, and White 2022), or to model spatial-temporal graph data (Jiang et al 2022;Lou et al 2021). In intra-graph FGL, instead of complete graphs, each client only owns a subgraph of the entire graph and the learning scheme is to deal with the missing links (Chen et al 2021), i.e., generating missing neighbors (Zhang et al 2021b), community discover (Baek et al 2022).…”

Section: Related Workmentioning

confidence: 99%

Federated Learning on Non-IID Graphs via Structural Knowledge Sharing

Tan

Liu

Long

et al. 2023

AAAI

View full text Add to dashboard Cite

Graph neural networks (GNNs) have shown their superiority in modeling graph data. Owing to the advantages of federated learning, federated graph learning (FGL) enables clients to train strong GNN models in a distributed manner without sharing their private data. A core challenge in federated systems is the non-IID problem, which also widely exists in real-world graph data. For example, local data of clients may come from diverse datasets or even domains, e.g., social networks and molecules, increasing the difficulty for FGL methods to capture commonly shared knowledge and learn a generalized encoder. From real-world graph datasets, we observe that some structural properties are shared by various domains, presenting great potential for sharing structural knowledge in FGL. Inspired by this, we propose FedStar, an FGL framework that extracts and shares the common underlying structure information for inter-graph federated learning tasks. To explicitly extract the structure information rather than encoding them along with the node features, we define structure embeddings and encode them with an independent structure encoder. Then, the structure encoder is shared across clients while the feature-based knowledge is learned in a personalized way, making FedStar capable of capturing more structure-based domain-invariant information and avoiding feature misalignment issues. We perform extensive experiments over both cross-dataset and cross-domain non-IID FGL settings, demonstrating the superiority of FedStar.

show abstract

“…Federated learning (FL) represents a crucial machine learning paradigm in which distributed clients (e.g., several medical institutions) collaboratively train a shared global model while retaining their private data. 10 , 11 , 12 , 13 However, inherent biases may arise in the federated model because of spurious correlations and distribution shifts across data subpopulations. 14 , 15 , 16 , 17 , 18 Consequently, the model’s performance may significantly degrade for certain data subpopulations, leading to concerns regarding unfairness, particularly in critical domains such as healthcare.…”

Section: Introductionmentioning

confidence: 99%

Unified fair federated learning for digital healthcare

Zhang,

Shuai,

Kuang

et al. 2024

Patterns

View full text Add to dashboard Cite

Federated learning of molecular properties with graph neural networks in a heterogeneous setting

Cited by 17 publications

References 48 publications

Transforming organic chemistry research paradigms: Moving from manual efforts to the intersection of automation and artificial intelligence

Transforming organic chemistry research paradigms: Moving from manual efforts to the intersection of automation and artificial intelligence

Federated Learning on Non-IID Graphs via Structural Knowledge Sharing

Unified fair federated learning for digital healthcare

Contact Info

Product

Resources

About