Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore the comprehensive Stanford course that covers the structure and analysis of large social and information networks through machine learning approaches. Master traditional feature-based methods for nodes, links, and graphs before diving into modern node embeddings and random walk approaches. Learn fundamental algorithms like PageRank, matrix factorization, and collective classification techniques for network analysis. Develop expertise in Graph Neural Networks (GNNs), starting with basic deep learning concepts and progressing to advanced GNN architectures, training methods, and expressiveness analysis. Investigate heterogeneous networks and knowledge graph embeddings, including completion algorithms and reasoning techniques like Query2box. Study neural subgraph matching, community detection algorithms including Louvain, and both classical and deep generative models for graphs. Address practical challenges such as GNN limitations, scaling techniques like GraphSAGE and Cluster GCN, and specialized applications in computational biology. Conclude with advanced topics including pre-training strategies, hyperbolic embeddings, and comprehensive design principles for Graph Neural Networks, providing both theoretical foundations and practical skills for analyzing large-scale network data.
