Neural Networks in Python from Scratch: Complete guide (Udemy.com)

The course is very clear to detail and teaches you from beginner to advanced. The only problem I got was in loading some library of pybrain where there was an error connected to scipy. According to the course, the error was in removing 2 from em2 but I think the libraries now available on internet have a different form and the error got is related to scipy

This is one of the better Udemy courses out there. The instructor doesn't assume any prior knowledge, and each line of code is explained in detail. I think the course does a good job of starting with very simple concepts and then building from there. My one critique is that some of the functions used are deprecated, so you'll have to spend a lot of time dealing with error messages and searching for workarounds to actually apply the models.

The course was nice and provided a good introduction into Deep Learning and Neural Networks. However, during Section 2 of the course when we were studying about the derivations of the networks, I believe that there were a few mistakes. First of all, the formula for the Sigmoid activation function is not (1/(1-e^-(w*x +b)), but rather (1/(1+e^-(w*x +b)). Secondly, the calculations do not make sense when the Loss function is set to y - y_hat. In order to obtain the same results I had to use (y-y_hat)^2, otherwise when I was deriving the calculation were incorrect. On a separate note, I liked that a list of books was provided. I will definitely will be picking those up for more in depth mathematical overview of NN. Thank you!

Yes - this really is a good course. I would not have got this kind of perspective working on my own. The way it builds from the implementation of Perceptrons and multilayer neural networks is particularly great - showing us all there is no "Magic" to neural networks. I also enjoyed being introduced to the "real life" environments such as TensorFlow Well Done! Tim Schofield

Good explanations and definitely helpful for understanding and learning to develop neural networks. Only downside is that it can be a little slow and some aspects can feel a little repetitive by the end of the course.

I took course until the end and then writed review. Good explanation of terms and concepts. Easy to understand complicated termins and concepts. Understandable English pronunciation and at the right noreal speed. Code examples without mistakes, I tried in my Intellij IDEA environment.

This is very good. The explanations are very thorough. The HWs require a little additional knowledge, but the solutions are explained very well. If you work in a different environment (e.g., Anaconda), expect small changes in the code to be required. That may be due to working with newer versions of the libraries. Overall, a very good experience.

I thoroughly enjoyed Jones' course. The early lectures involved adjusting to some pronunciation quirks because English is not his native language or mather tongue I assume, but this was no big deal and any mispronunciations were contextually apparent anyway. The material covered stands as a good introduction. Will you walk into a job on Neural networks from this course alone? Of course not. Will tis course set you up to tackle a more advanced course or advanced book? Yes, I believe it would. Would I take another course with Jones? Yes. I liked the informative, no nonsense and practical focused approach. Worth the 4 stars. Thank you, Jones.

I had a lot of theoretical knowledge about neural networks in college but I was never able to understand the math or how to go about it in code. This quick course gave me confidence in this exciting field and I am now pursuing knowledge about deep learning. This course is very clear with its explanations and the homeworks are a great way to practice. Being taught the basics of industry standard libraries such as TensorFlow and PyTorch is such a benefit.

This was a really good first course in ANN! You do need some python under your belt and if you come form a scientific background it will help understand everything. I enjoyed learning the ANN from scratch as I always pictured it in my mind when going through the rest of the course. Definitely recommended!!