Stanford CS224W: Machine Learning with Graphs | 2021 | Lecture 13.3 - Louvain Algorithm

For more information about Stanford’s Artificial Intelligence professional and graduate programs, visit: https://stanford.io/3jQEhdD

Jure Leskovec
Computer Science, PhD

In this part, we introduce the Louvain Algorithm which is a popular fast and scalable method with O(nlongn) runtime to detect communities in a graph. This algorithm is widely applicable and can be used with weighted graphs and for finding heirarchable communities. You will see Louvain algorithm works greedily to maximize modularity operating in two phases. In phase I, local changes can be made to node-community memberships. And in phase II, identified communities are aggregated into super-nodes to form a new network. We see how this process can be repeated to convergence to get the final communities.

To follow along with the course schedule and syllabus, visit:
http://web.stanford.edu/class/cs224w/

#machinelearning #machinelearningcourse