Graph Neural Network (Udemy.com)

This course presents very well the basic concepts of Graph Neural Networks with some practical sessions. However, I would have given 5 stars if there was parts about Graph Neural Networks for time series processing, and some workshops on their implementation using PyTorch.

Great course to understand GNN. I like how the instructor arranged the course and the split between theory and practice example. I have a new project to work on related to GNN and before the course i do not know what is GNN is to begin with but now i can work on the project with a comprehensive understanding. Still requires more reading if you want to dig deeper into the subject

It wasn't a good match for me. I was looking for a course with exercises, quizzes, and challenges. Also, This course was very very brief and didn't include applications of GNN. It didn't include edge classification and many other types of GNNs. I believe the name of course should change to something less general. To be fair, I should mention that the theoretical parts were explained very well and could become better with better representation.

This is a great introduction into the topic of graph neural networks. What I like about the course is that the author mixes theory sessions where he walks through the basic concept of papers with practical sessions where he shows how the theory can be implemented with PyTorch. At the end of the course you have a solid understanding of the main concepts behind graph neural networks.

As it happens with learning any new topic, I still needed to look at multiple sources (such as, youtube videos, blogs), but this course is a good singular resource that covers the basics. I found essential information here that I had not found elsewhere. I liked how the instructor linked one topic when moving on to another, which was helpful in understanding the whole GNN picture.

The overall flow is great. The content is good and well explained. I am glad I took this course. Additionally, I would also like to know how to populate the initial graph with data after predicting e.g. labels (and show how to work with it).

Very good depth of explanation. This course was enough to help me use my understanding non-graph neural network approaches and graph data structures, and ultimately come to a pretty solid understanding of how to start translating things that are best represented in graph structures into structures that existing ML tools can work with.

The course is really good, it was like a research story of how the concept of graphs came into Ai! I now understand how the research go from 2007, from the idea of graphs that was pitched, till the recent years when graphs could finally be incorporated into neural networks that went through so many conundrums. Inspiring to know the zeal behind the research, to not to give up until a solution was found! Amazing..

I enjoyed the course and learned a lot from it. The content is well organized and focused on practical application of Graph ML models. I especially likes and learned a lot about getting unseen nodes predictions using Transductive Learning.

The course provides a broad-based superficial overview of Graph Neural Networks. The effort to include prior methods, such as embeddings is commendable. Materials on YouTube have supplemented the course greatly. Overall, however, the details and lack of lecture slides for GNN is apparent and not surmountable to a more than $10 cost on Udemy.