Dear Mr. Foster,
I am currently studying for an M.S. in Artificial Intelligence, and I have been using your book Generative Deep Learning as part of my course learning. I wanted to share some feedback from the perspective of a serious learner who is trying to understand the material deeply, not just read it quickly.
I find the book valuable, and I appreciate the effort behind explaining different generative model families. However, I often feel lost when moving from one chapter to the next because the connections between chapters are not always made explicit.
For example, after studying earlier generative models, Chapter 6 introduces normalizing flows, which feel like a completely new world. I understand that each chapter may cover a different family of generative models, but I think learners would benefit from stronger bridge sections explaining how the new chapter connects to the previous models, what problem the new model family solves, and why we are studying it at that point in the book.
A short transition, such as “GANs generate samples but do not provide exact likelihood, while normalizing flows are introduced because they provide invertible transformations and explicit density estimation,” would help readers understand the purpose of the chapter before entering the mathematical details.
My feedback is mainly that clearer chapter-to-chapter transitions would make the learning path much easier to follow, especially for students who are building their understanding step by step.
Thank you for the book and for the work you have put into helping learners study generative deep learning.
Sincerely,
Basil
Dear Mr. Foster,
I am currently studying for an M.S. in Artificial Intelligence, and I have been using your book Generative Deep Learning as part of my course learning. I wanted to share some feedback from the perspective of a serious learner who is trying to understand the material deeply, not just read it quickly.
I find the book valuable, and I appreciate the effort behind explaining different generative model families. However, I often feel lost when moving from one chapter to the next because the connections between chapters are not always made explicit.
For example, after studying earlier generative models, Chapter 6 introduces normalizing flows, which feel like a completely new world. I understand that each chapter may cover a different family of generative models, but I think learners would benefit from stronger bridge sections explaining how the new chapter connects to the previous models, what problem the new model family solves, and why we are studying it at that point in the book.
A short transition, such as “GANs generate samples but do not provide exact likelihood, while normalizing flows are introduced because they provide invertible transformations and explicit density estimation,” would help readers understand the purpose of the chapter before entering the mathematical details.
My feedback is mainly that clearer chapter-to-chapter transitions would make the learning path much easier to follow, especially for students who are building their understanding step by step.
Thank you for the book and for the work you have put into helping learners study generative deep learning.
Sincerely,
Basil