ICLR 2021 Keynote - Geometric Deep Learning: The Erlangen Programme of ML - M Bronstein Michael Bronstein 2021-06-08 | Geometric Deep Learning: The Erlangen Programme of ML - ICLR 2021 Keynote by Michael Bronstein (Imperial College London / IDSIA / Twitter)“Symmetry, as wide or as narrow as you may define its meaning, is one idea by which man through the ages has tried to comprehend and create order, beauty, and perfection.” This poetic definition comes from the great mathematician Hermann Weyl, credited with laying the foundation of our modern theory of the universe. Another great physicist, Philip Anderson, said that "it is only slightly overstating the case to say that physics is the study of symmetry."In mathematics, symmetry was crucial in the foundation of geometry as we know it in the 19th century. Now it could have a similar impact on another emerging field. Deep Learning success in recent decades is significant – from revolutionising data science to landmark achievements in computer vision, board games, and protein folding. At the same time, a lack of unifying principles makes it is difficult to understand the relations between different neural network architectures resulting in the reinvention and re-branding of the same concepts. Michael Bronstein is a professor at Imperial College London and Head of Graph ML Research at Twitter, who is working to bring geometric unification of deep learning through the lens of symmetry. In his ICLR 2021 keynote lecture, he presents a common mathematical framework to study the most successful network architectures, giving a constructive procedure to build future machine learning in a principled way that could be applied in new domains such as social science, biology, and drug design. Based on M. M. Bronstein, J. Bruna, T. Cohen, P. Veličković, Geometric Deep Learning: Grids, Groups, Graphs, Geodesics, and Gauges, arXiv:2104.13478, 2021 (arxiv.org/abs/2104.13478)Accompanying blog post: towardsdatascience.com/geometric-foundations-of-deep-learning-94cdd45b451d?sk=184532175cb936d7b25d9adebd512629More information: geometricdeeplearning.comAnimation: Jakub Makowski
Lecture 3: Sheaf Neural Networks - Cristian Bodnar Michael Bronstein 2022-08-09 | Video recording of the First Italian Summer School on Geometric Deep Learning, which took place in July 2022 in Pescara. Slides: https://www.sci.unich.it/geodeep2022/slides/GDL_SummerSchool_Part2.pdfBlog: towardsdatascience.com/graph-neural-networks-beyond-weisfeiler-lehman-and-vanilla-message-passing-bc8605fa59a?sk=3a058b427798d601ebb2d618fca89a8f
Prerequisites II: Topology - Cristian Bodnar Michael Bronstein 2022-08-09 | Video recording of the First Italian Summer School on Geometric Deep Learning, which took place in July 2022 in Pescara. Slides: https://www.sci.unich.it/geodeep2022/slides/GDL_SummerSchool_Part0.pdf
Lecture 6: Gauge-equivariant Mesh CNN - Pim de Haan Michael Bronstein 2022-08-09 | Video recording of the First Italian School on Geometric Deep Learning held in Pescara in July 2022. Slide: https://www.sci.unich.it/geodeep2022/slides/2022-07-27%20Mesh%20-%20First%20Italian%20GDL%20School.pdf
Lecture 7: From Equivariance to Naturality - Pim de Haan Michael Bronstein 2022-08-09 | Video recording of the First Italian School on Geometric Deep Learning held in Pescara in July 2022. Slides: https://www.sci.unich.it/geodeep2022/slides/2022-07-27%20Naturality%20@%20First%20Italian%20GDL%20Summer%20School.pdf
Lecture 5: Equivariant CNNs II (Riemannian manifolds) - Maurice Weiler Michael Bronstein 2022-08-09 | Video recording of the First Italian School on Geometric Deep Learning held in Pescara in July 2022. Slides: https://www.sci.unich.it/geodeep2022/slides/CoordinateIndependentCNNs.pdf
Prerequisites IV: Category Theory - Pim de Haan Michael Bronstein 2022-08-09 | Video recording of the First Italian Summer School on Geometric Deep Learning, which took place in July 2022 in Pescara. Slides: https://www.sci.unich.it/geodeep2022/slides/2022-07-25%20Intro%20to%20categories.pptx
Prerequisites III: Manifolds & Fiber Bundles - Maurice Weiler Michael Bronstein 2022-08-09 | Video recording of the First Italian Summer School on Geometric Deep Learning, which took place in July 2022 in Pescara. Slides: https://www.sci.unich.it/geodeep2022/slides/Manifolds_and_Fiber_Bundles.pdf
Lecture 4: Equivariant CNNs I (Euclidean Spaces) - Maurice Weiler Michael Bronstein 2022-08-09 | Video recording of the First Italian School on Geometric Deep Learning held in Pescara in July 2022. Slides: https://www.sci.unich.it/geodeep2022/slides/GroupEquivariantConvolutionalNetworksOnEuclideanSpaces.pdf
Lecture 8: Curvature & Oversquashing in GNNs - Francesco Di Giovanni Michael Bronstein 2022-08-09 | Video recording of the First Italian School on Geometric Deep Learning held in Pescara in July 2022.Slides: https://www.sci.unich.it/geodeep2022/slides/GRAFF_presentation%20(18).pdfBlog post: towardsdatascience.com/over-squashing-bottlenecks-and-graph-ricci-curvature-c238b7169e16?sk=f5cf01cbd57b4fee8fb3a89d447e56a9
Prerequisites I: Groups, Representations & Equivariance - Maurice Weiler Michael Bronstein 2022-08-09 | Video recording of the First Italian Summer School on Geometric Deep Learning, which took place in July 2022 in Pescara. Slides: https://www.sci.unich.it/geodeep2022/slides/Groups_Representations_and_Equivariance.pdf
Lecture 9: GNNs as Dynamic Systems - Francesco Di Giovanni Michael Bronstein 2022-08-09 | Video recording of the First Italian School on Geometric Deep Learning held in Pescara in July 2022. Slides: https://www.sci.unich.it/geodeep2022/slides/GRAFF_presentation%20(17).pdf
Lecture 10: Whats Next? - Michael Bronstein Michael Bronstein 2022-08-09 | Video recording of the First Italian School on Geometric Deep Learning held in Pescara in July 2022. Slides: https://www.sci.unich.it/geodeep2022/slides/Pescara%202022%20-%20conclusions.pdfBlog post: towardsdatascience.com/graph-neural-networks-beyond-weisfeiler-lehman-and-vanilla-message-passing-bc8605fa59a?sk=3a058b427798d601ebb2d618fca89a8f
Lecture 1: A Brief History of Geometric Deep Learning - Michael Bronstein Michael Bronstein 2022-08-09 | Video recording of the First Italian Summer School on Geometric Deep Learning, which took place in July 2022 in Pescara. Slides: https://www.sci.unich.it/geodeep2022/slides/Pescara%202022%20-%20Intro.pdf
Lecture 2: Topological Message Passing - Cristian Bodnar Michael Bronstein 2022-08-09 | Video recording of the First Italian Summer School on Geometric Deep Learning, which took place in July 2022 in Pescara. Slides: https://www.sci.unich.it/geodeep2022/slides/GDL_SummerSchool_Part1.pdfBlog post: towardsdatascience.com/a-new-computational-fabric-for-graph-neural-networks-280ea7e3ed1a?sk=3d4185067ef3c1cf81793deada08e18fRelated papers: papers.nips.cc/paper/2021/file/157792e4abb490f99dbd738483e0d2d4-Paper.pdfhttp://proceedings.mlr.press/v139/bodnar21a/bodnar21a.pdf
AMMI 2022 Course Geometric Deep Learning - Seminar 1 (Physics-based GNNs) - Francesco Di Giovanni Michael Bronstein 2022-07-27 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July 2022 Seminar 1 - Graph neural networks through the lens of multi-particle dynamics and gradient flows - Francesco Di Giovanni (Twitter)Slides: dropbox.com/s/ecowz78xphdhtw0/AIMS%202022%20-%20Seminar%201%20-%20Physics-based%20GNNs.pdf?dl=0
AMMI 2022 Course Geometric Deep Learning - Seminar 5 (AlphaFold) - Russ Bates Michael Bronstein 2022-07-27 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July 2022 Seminar 5 - Highly accurate protein structure prediction with AlphaFold - Russ Bates (DeepMind)Slides: dropbox.com/s/lgu6658b7kv2s9w/AIMS%202022%20-%20Seminar%205%20-%20AlphaFold.pdf?dl=0
AMMI 2022 Course Geometric Deep Learning - Lecture 4 (Geometric Priors II) - Joan Bruna Michael Bronstein 2022-07-27 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July 2022 by Michael Bronstein (Oxford), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July 2022 by Michael Bronstein (Oxford), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 4: Invariant function classes • Learning under invariance • Compositionality • Multiresolution analysis • Scale separation • Combining invariance and Scale separationSlides: dropbox.com/s/6ueh81pfvpp8vx0/AIMS%202022%20-%20Lecture%204%20-%20geometric%20priors%20II.pdf?dl=0Additional materials: www.geometricdeeplearning.com
AMMI 2022 Course Geometric Deep Learning - Lecture 6 (Graphs & Sets II) - Petar Veličković Michael Bronstein 2022-07-27 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July 2022 by Michael Bronstein (Oxford), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 6: General attributed graphs • Graph networks • DeepSets • Transformers • Neural relational inference • Dynamic Graph CNN • Differentiable Graph Module • Pointer Graph Networks • Weisfeiler-Lehman test • Higher-order GNNsSlides: dropbox.com/s/itsbi113f7ogpbv/AIMS%202022%20-%20Lecture%206%20-%20Graphs%20%26%20Sets%20II.pdf?dl=0Additional materials: www.geometricdeeplearning.com
AMMI 2022 Course Geometric Deep Learning - Lecture 7 (Grids) - Joan Bruna Michael Bronstein 2022-07-27 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July 2022 by Michael Bronstein (Oxford), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 1: Symmetry through the centuries • First neural networks and the “Perceptron affair” • The curse of dimensionality • First geometric architectures: neocognitron and CNNs • Chemical precursors of GNNs • Geometric deep learning blueprint • The "5G" of Geometric deep learning • Course outlineSlides: Lecture 7: Grids and Translations • Translation group • Shift operator • Linear invariants and equivariants • Fourier transform • Convolution • Fourier invariants • Deformation stability • Multiscale representations • Wavelets • Scattering • CNNsSlides: dropbox.com/s/xjwb6weegz9vuwz/AIMS%202022%20-%20Lecture%207%20-%20Grids.pdf?dl=0Blog post: towardsdatascience.com/deriving-convolution-from-first-principles-4ff124888028?sk=0d77e2fd7863d457aeb2dac620dd133cAdditional materials: www.geometricdeeplearning.com
AMMI 2022 Course Geometric Deep Learning - Seminar 2 (Subgraph GNNs) - Fabrizio Frasca Michael Bronstein 2022-07-27 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July 2022 Seminar 2 - Subgraphs for more powerful GNNs - Fabrizio Frasca (Twitter)Slides: dropbox.com/s/tnuhppf1fqmv6y9/AIMS%202020%20-%20Seminar%202%20-%20Subgraph%20GNNs.pdf?dl=0Blog post: towardsdatascience.com/using-subgraphs-for-more-expressive-gnns-8d06418d5ab?sk=8806ffcd9ecf74c440d40df53528c1c7
AMMI 2022 Course Geometric Deep Learning - Lecture 8 (Groups & Homogeneous spaces) - Taco Cohen Michael Bronstein 2022-07-27 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July 2022 by Michael Bronstein (Oxford), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 8: Group convolution • Regular representation • Spherical CNNs • Homogeneous spaces • Cosets • Quotients • Coset pooling • Scalar, Regular, and Induced representation • Steerable CNNs • Kernel constraintSlides: dropbox.com/s/869fn3uuwt8whdo/AIMS%202022%20-%20Lecture%208%20-%20Groups.pdf?dl=0Additional materials: www.geometricdeeplearning.com
AMMI 2022 Course Geometric Deep Learning - Seminar 4 (Neural Sheaf Diffusion) - Cristian Bodnar Michael Bronstein 2022-07-27 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July 2022 Seminar 5 - Neural Sheaf Diffusion - Cristian Bodnar (Cambridge)Slides: dropbox.com/s/7hyqlwh45bdkv0f/AIMS%202022%20-%20Seminar%204%20-%20Neural%20sheaf%20diffusion.pdf?dl=0Blog post: towardsdatascience.com/neural-sheaf-diffusion-for-deep-learning-on-graphs-bfa200e6afa6?sk=0b2f814a1180a64460699f6a3277053a
AMMI 2022 Course Geometric Deep Learning - Seminar 3 (Equivariance in ML) - Geordie Williamson Michael Bronstein 2022-07-27 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July 2022 Seminar 3 - Equivariance in Machine Learning - Geordie Williamson (U Sydney)Slides: dropbox.com/s/w10mfq072zkhsap/AIMS%202022%20-%20Seminar%203%20-%20Equivariance%20in%20ML.pdf?dl=0
AMMI 2022 Course Geometric Deep Learning - Lecture 9 (Manifolds) - Michael Bronstein Michael Bronstein 2022-07-27 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July 2022 by Michael Bronstein (Oxford), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 9: Euclidean vs Non-Euclidean convolution • Manifolds • Tangent vectors • Riemannian metric • Geodesics • Parallel transport • Exponential map • Convolution on manifolds • Domain deformation • Isometries • Deformation invariance • Scalar and vector fields • Gradient, Divergence, and Laplacian operators • Heat and Wave equations • Manifold Fourier transform • Spectral convolution • Meshes • Discrete Laplacians • ChebNet Slides: dropbox.com/s/8pc7b53z0w2ui15/AIMS%202022%20-%20Lecture%209%20-%20maniflds%2C%20meshes%2C%20and%20geometric%20graphs.pdf?dl=0Additional materials: www.geometricdeeplearning.com
AMMI 2022 Course Geometric Deep Learning - Lecture 2 (Learning in High Dimensions) - Joan Bruna Michael Bronstein 2022-07-27 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July 2022 by Michael Bronstein (Oxford), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 2: Basic notions in learning • Challenges of learning in high dimension • Learning Lipschitz functions • Universal approximationSlides: dropbox.com/s/3hp6utrkm3d6ak6/AIMS%202%20-%20learning%20in%20high%20dimensions%202022.pdf?dl=0
AMMI 2022 Course Geometric Deep Learning - Lecture 3 (Geometric Priors I) - Taco Cohen Michael Bronstein 2022-07-27 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July 2022 by Michael Bronstein (Oxford), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 3: Symmetries • Abstract groups • Symmetry groups • Group actions • Group representations • Invariance and Equivariance Slides: dropbox.com/s/ydxa5lqm1obb6yx/AIMS%202022%20-%20Lecture%203%20-%20Geometric%20Priors%20I.pdf?dl=0Additional materials: www.geometricdeeplearning.com
AMMI 2022 Course Geometric Deep Learning - Lecture 12 (Applications & Trends) - Michael Bronstein Michael Bronstein 2022-07-27 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July 2022 by Michael Bronstein (Oxford), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 12: What's next? • Beyond traditional Message Passing • Algorithmic Reasoning • Causal inferences • Knowledge Graphs • Data vs Computational graphs • Dynamic graphs • Generative models • Standardised benchmarks • Applications • Hardware beyond GPUs • Recommender systems • Fake news detection • Self-driving cars • 3D avatars • High-energy physics • Structural biology and protein science • Chemistry and drug design • Drug repositioning • Hyperfoods • Physics-inspired learning on graphs • Neural diffusion equations • Graph Beltrami flow • Ricci curvature & Bottlenecks • Graph rewiring • Sheaf diffusion • Graph coupled oscillatorsSlides: dropbox.com/s/9zcrcohird72ulv/AIMS%202022%20-%20Lecture%2012%20-%20conclusions.pdf?dl=0Additional materials: www.geometricdeeplearning.com
AMMI 2022 Course Geometric Deep Learning - Lecture 5 (Graphs & Sets) - Petar Veličković Michael Bronstein 2022-07-27 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July 2022 by Michael Bronstein (Oxford), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 5: Learning on sets • Permutations • Permutation invariance • DeepSets • Permutation equivariance • General blueprint of learning on sets • Learning on graphs • Locality • General blueprint of learning on graphs • Three flavors of GNNs • Convolutional GNNs • Attentional GNNs • Message Passing GNNsSlides: dropbox.com/s/fola7dmqtjfiwqr/AIMS%202022%20-%20Lecture%205%20-%20Graphs%20%26%20Sets%20I.pdf?dl=0Additional materials: www.geometricdeeplearning.com
AMMI 2022 Course Geometric Deep Learning - Lecture 11 (Beyond Groups) - Petar Veličković Michael Bronstein 2022-07-27 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July 2022 by Michael Bronstein (Oxford), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 11: Category Theory • Set category • Functors • Natural transformations • Natural graph networksSlides: dropbox.com/s/4nt0fqrrhw2yb63/AIMS%202022%20-%20Lecture%2011%20-%20Beyond%20Groups.pdf?dl=0Additional materials: www.geometricdeeplearning.com
AMMI 2022 Course Geometric Deep Learning - Lecture 10 (Gauges) - Taco Cohen Michael Bronstein 2022-07-27 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July 2022 by Michael Bronstein (Oxford), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 10: Gauges • Gauge transformations • Examples of gauge symmetry • Weight sharing on manifolds • Geodesic convolution • Gauge-equivariant convolution • HexaConv • Gauge-equivariant Mesh CNNs • Fibre bundles • Principal bundles • Bundle maps • Universality resultsSlides: dropbox.com/s/fj4q3gv17xqpv0r/AIMS%202022%20-%20Lecture%2010%20-%20Gauges.pdf?dl=0Additional materials: www.geometricdeeplearning.com
AMMI 2022 Course Geometric Deep Learning - Lecture 1 (Introduction) - Michael Bronstein Michael Bronstein 2022-07-23 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July 2022 by Michael Bronstein (Oxford), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 1: Symmetry through the centuries • First neural networks and the “Perceptron affair” • The curse of dimensionality • First geometric architectures: neocognitron and CNNs • Chemical precursors of GNNs • Geometric deep learning blueprint • The "5G" of Geometric deep learning • Course outlineSlides: dropbox.com/s/v8725hkhbejpllw/AIMS%202022%20-%20Lecture%201%20-%20introduction.pdf?dl=0Additional materials: www.geometricdeeplearning.com
AMMI Course Geometric Deep Learning - Lecture 4 (Geometric Priors II) - Joan Bruna Michael Bronstein 2021-08-08 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July-August 2021 by Michael Bronstein (Imperial College/Twitter), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 4: Invariant function classes • Learning under invariance • Compositionality • Multiresolution analysis • Scale separation • Combining invariance and Scale separationSlides: bit.ly/37rWbgdAdditional materials: www.geometricdeeplearning.com
AMMI Course Geometric Deep Learning - Lecture 5 (Graphs & Sets I) - Petar Veličković Michael Bronstein 2021-08-08 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July-August 2021 by Michael Bronstein (Imperial College/Twitter), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 5: Learning on sets • Permutations • Permutation invariance • DeepSets • Permutation equivariance • General blueprint of learning on sets • Learning on graphs • Locality • General blueprint of learning on graphs • Three flavors of GNNs • Convolutional GNNs • Attentional GNNs • Message Passing GNNsSlides: bit.ly/2X75EHYAdditional materials: www.geometricdeeplearning.com
AMMI Course Geometric Deep Learning - Lecture 1 (Introduction) - Michael Bronstein Michael Bronstein 2021-08-08 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July-August 2021 by Michael Bronstein (Imperial College/Twitter), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 1: Symmetry through the centuries • The curse of dimensionality • Geometric priors • Invariance and equivariance • Geometric deep learning blueprint • The "5G" of Geometric deep learning • Graphs • Grids • Groups • Geodesics • Gauges • Course outlineSlides: bit.ly/3iw6AO9Additional materials: www.geometricdeeplearning.com
AMMI Course Geometric Deep Learning - Lecture 10 (Gauges) - Taco Cohen Michael Bronstein 2021-08-08 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July-August 2021 by Michael Bronstein (Imperial College/Twitter), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 10: Gauges • Gauge transformations • Examples of gauge symmetry • Weight sharing on manifolds • Geodesic convolution • Gauge-equivariant convolution • HexaConv • Gauge-equivariant Mesh CNNs • Fibre bundles • Principal bundles • Bundle maps • Universality resultsSlides: bit.ly/2VFC63bAdditional materials: www.geometricdeeplearning.com
AMMI Course Geometric Deep Learning - Lecture 3 (Geometric Priors I) - Taco Cohen Michael Bronstein 2021-08-08 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July-August 2021 by Michael Bronstein (Imperial College/Twitter), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 3: Symmetries • Abstract groups • Symmetry groups • Group actions • Group representations • Invariance and Equivariance Slides: bit.ly/3s1PACvAdditional materials: www.geometricdeeplearning.com
AMMI Course Geometric Deep Learning - Lecture 11 (Sequences & Time Warping) - Petar Veličković Michael Bronstein 2021-08-08 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July-August 2021 by Michael Bronstein (Imperial College/Twitter), Joan Bruna (NYU), Taco Cohen (Quallcom), and Petar Veličković (DeepMind)Lecture 11: Static and dynamic domains • Recurrent Neural Networks • Translation invariance • Time warping • Warping the ODE • Discrete Warped RNNs • Gated RNNs • LSTMSlides: bit.ly/3xBpUOdAdditional materials: www.geometricdeeplearning.com
AMMI Course Geometric Deep Learning - Lecture 2 (Learning in High Dimensions) - Joan Bruna Michael Bronstein 2021-08-08 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July-August 2021 by Michael Bronstein (Imperial College/Twitter), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 2: Basic notions in learning • Challenges of learning in high dimension • Learning Lipschitz functions • Universal approximationSlides: bit.ly/3yB5A14Additional materials: www.geometricdeeplearning.com
AMMI Course Geometric Deep Learning - Lecture 7 (Grids) - Joan Bruna Michael Bronstein 2021-08-08 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July-August 2021 by Michael Bronstein (Imperial College/Twitter), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 7: Grids and Translations • Translation group • Shift operator • Linear invariants and equivariants • Fourier transform • Convolution • Fourier invariants • Deformation stability • Multiscale representations • Wavelets • Scattering • CNNsSlides: bit.ly/3iu0GgrAdditional materials: www.geometricdeeplearning.com
AMMI Course Geometric Deep Learning - Lecture 9 (Manifolds & Meshes) - Michael Bronstein Michael Bronstein 2021-08-08 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July-August 2021 by Michael Bronstein (Imperial College/Twitter), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 9: Euclidean vs Non-Euclidean convolution • Manifolds • Tangent vectors • Riemannian metric • Geodesics • Parallel transport • Exponential map • Convolution on manifolds • Domain deformation • Pushfowards and Pullback • Isometries • Deformation invariance • Scalar and vector fields • Gradient, Divergence, and Laplacian operators • Heat and Wave equations • Manifold Fourier transform • Spectral convolution • Meshes • Discrete Laplacians • ChebNet • Graph Convolutional Network • sGCN • SIGN Slides: bit.ly/3lFTobtAdditional materials: www.geometricdeeplearning.com
AMMI Course Geometric Deep Learning - Lecture 12 (Applications & Conclusions) - Michael Bronstein Michael Bronstein 2021-08-08 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July-August 2021 by Michael Bronstein (Imperial College/Twitter), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 12: What's next? • Beyond Message Passing • Algorithmic Reasoning • Causal inferences • Knowledge Graphs • Data vs Computational graphs • Neural PDEs • Dynamic graphs • Generative models • Standardised benchmarks • Applications • Hardware beyond GPUs • Recommender systems • Fake news detection • Self-driving cars • 3D avatars • Face from DNA • Applications in high-energy physics • Applications in structural biology and protein science • AlphaFold 2 • MaSIF • Applications in chemistry and drug design • Virtual drug screening • Drug repositioning • HyperfoodsSlides: dropbox.com/s/9zcrcohird72ulv/AIMS%202022%20-%20Lecture%2012%20-%20conclusions.pdf?dl=0Additional materials: www.geometricdeeplearning.com
AMMI Course Geometric Deep Learning - Lecture 6 (Graphs & Sets II) - Petar Veličković Michael Bronstein 2021-08-08 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July-August 2021 by Michael Bronstein (Imperial College/Twitter), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 6: General attributed graphs • Graph networks • DeepSets • Transformers • Neural relational inference • Dynamic Graph CNN • Differentiable Graph Module • Pointer Graph Networks • Weisfeiler-Lehman test • Higher-order GNNsSlides: bit.ly/37pYssrAdditional materials: www.geometricdeeplearning.com
AMMI Course Geometric Deep Learning - Lecture 8 (Groups & Homogeneous spaces) - Taco Cohen Michael Bronstein 2021-08-08 | Video recording of the course "Geometric Deep Learning" taught in the African Master in Machine Intelligence in July-August 2021 by Michael Bronstein (Imperial College/Twitter), Joan Bruna (NYU), Taco Cohen (Qualcomm), and Petar Veličković (DeepMind)Lecture 8: Group convolution • Regular representation • Spherical CNNs • Homogeneous spaces • Cosets • Quotients • Coset pooling • Scalar, Regular, and Induced representation • Steerable CNNs • Kernel constraintSlides: bit.ly/3AjUyxuAdditional materials: www.geometricdeeplearning.com
