Sorry, this is a bad video. I gave a thumbs down. Part 2 is bad as well. It's phony, basically because you don't discuss the key concepts at all. It is just 'how wonderful they are' and 'great stuff they can do', not actually how they work. They second video shows some fancy formulas but also never get to a good conceptual explanation. I'm going to put a dislike on that one too. Please do better or don't take our time. No offense.
Thanks. I still can't grasp clearly the intuition of how it can be different from a fully connected layers network (FCN). Can we consider FCN with a sufficient hidden layer as a generalized form of GCN, where GCN is more constrained, thus allowing for better learning in graph-structured problems?
Sorry, this is a bad video. I gave a thumbs down. Part 2 is bad as well. It's phony, basically because you don't discuss the key concepts at all. It is just 'how wonderful they are' and 'great stuff they can do', not actually how they work. They second video shows some fancy formulas but also never get to a good conceptual explanation. I'm going to put a dislike on that one too.
Please do better or don't take our time. No offense.
Hi Jacob, thank you for this video. Really interested, please make a video about a basic mathematics of GNN
Succinct but well articulated deep insights on Graphs, GML, GNNs - great work. Thank you Jacob.
Thanks. I still can't grasp clearly the intuition of how it can be different from a fully connected layers network (FCN). Can we consider FCN with a sufficient hidden layer as a generalized form of GCN, where GCN is more constrained, thus allowing for better learning in graph-structured problems?
Hi Jacob! Thank you so much for this video. Would sharely share it with my colleagues.
omg, you're the best, your tutorial was a life saver
keep going, you’re the best in term of making GNN simpler
great intro to GNN much better than other videos.
Great video! You convinced me to study GNN
Like it !!! Very much