Machine Learning for Graphs and Sequential Data (MLGS)

Below you can find the videos of our lecture "Machine Learning for Graphs and Sequential Data" (MLGS). The videos for "Advanced Machine Learning: Deep Generative Models" can be found here.

Note: The videos presented here are recordings from previous years and are additional to the actual lecture. Therefore, the content in these videos can differ from the lecture and is not exhaustive.

If you spot any typos or mistakes, please send an e-mail to Prof. Günnemann with the subject [MLGS Videos]. Thank you!

• Introduction: Video
• Sequential Data
  • Autoregressive Models: Video
  • Markov Chains: Video
  • Hidden Markov Models (HMMs): Part1, Part2, Part3
  • Word Vectors/Embeddings: Video
  • Recurrent Neural Networks (e.g. RNN, LSTM): Video
  • Non-Recurrent Neural Networks (e.g. ConvNets, Transformer): Video
  • Temporal Point Processes: Part 1, Part 2
• Graphs
  • Introduction: Video
  • Generative Models: Part1, Part2
  • Clustering (e.g. Cuts, Graph Laplacian, Spectral Embedding/Clustering, SBM): Part1, Part2, Part3
  • Node Embeddings: Video
  • Ranking (e.g. PageRank): Video
  • Semi-Supervised Learning
    • Introduction, Concepts, Label Propagation: Video
    • Graph Neural Networks (GNNs): Video
  • Limitations of Graph Neural Networks: Part1, Part2, Part3
• Robustness
  • Introduction: Video
  • Adversarial Example Construction: Video
  • Improving Robustness: Video
  • Certifiable Robustness
    • Exact Certification: Video
    • Convex Relaxation: Video
    • Lipschitz Continuity: Video
    • Randomized Smoothing: Video
• The End: Video
• Recap Bayesian Networks: Video