Chevron Left
Back to Improving Deep Neural Networks: Hyperparameter Tuning, Regularization and Optimization

Learner Reviews & Feedback for Improving Deep Neural Networks: Hyperparameter Tuning, Regularization and Optimization by DeepLearning.AI

4.9
stars
62,825 ratings

About the Course

In the second course of the Deep Learning Specialization, you will open the deep learning black box to understand the processes that drive performance and generate good results systematically. By the end, you will learn the best practices to train and develop test sets and analyze bias/variance for building deep learning applications; be able to use standard neural network techniques such as initialization, L2 and dropout regularization, hyperparameter tuning, batch normalization, and gradient checking; implement and apply a variety of optimization algorithms, such as mini-batch gradient descent, Momentum, RMSprop and Adam, and check for their convergence; and implement a neural network in TensorFlow. The Deep Learning Specialization is our foundational program that will help you understand the capabilities, challenges, and consequences of deep learning and prepare you to participate in the development of leading-edge AI technology. It provides a pathway for you to gain the knowledge and skills to apply machine learning to your work, level up your technical career, and take the definitive step in the world of AI....

Top reviews

XG

Oct 30, 2017

Thank you Andrew!! I know start to use Tensorflow, however, this tool is not well for a research goal. Maybe, pytorch could be considered in the future!! And let us know how to use pytorch in Windows.

JS

Apr 4, 2021

Fantastic course and although it guides you through the course (and may feel less challenging to some) it provides all the building blocks for you to latter apply them to your own interesting project.

Filter by:

6151 - 6175 of 7,216 Reviews for Improving Deep Neural Networks: Hyperparameter Tuning, Regularization and Optimization

By Andrei M

•

Apr 1, 2018

I feel like the assignments involve a lot of cut and pasting of functions verbatim. It's a good start, but I'd like to go further and be challenged to solve a problem, rather than fill in the blanks. For example, try different optimizers to reach a particular learning speed on a given data set.

By Daniel M

•

Mar 8, 2021

Hands-on projects are very easy. The only ask for copying and pasting some lines of code. In contrast, questionnaires are well designed and reprepresent the knowledge acquired. Another course drawback is that the last programming practice employs TensorFlow 1, the version 2 was released in 2019.

By ehsan

•

Jun 16, 2020

Everything was fine but there were some issues that was not pleasant to me. like seeing that there's problemmes with the videos and they're not corrected and instead there's an extra reading about that. Or when Tensorflow 2 has come, I was expecting that the course also introduces newer version.

By Parham A

•

Aug 13, 2020

The course and the instructor are amazing, but I fell behind schedule by one week and the last assignment was locked, and when attempting to reset the deadline, a message saying "something went wrong" would pop up. The help center responded very quickly and professionally, and solved the issue.

By Dhruv S

•

Jun 8, 2020

This course I believe is one of the most vital one after the first course in the specialization! Professor Ng covers all the concepts required for you to understand and master this course.

You might have to refer to additional resources to get a complete grasp of the concept post each video.

By Eemeli L

•

Nov 19, 2019

Great and easy-to-follow introduction to improving deep neural networks. If you are already familiar with vector algebra, many things are explained quite slowly. One star left out because the content has not been polished, but there are minor errors here and there with separate corrections.

By Krishna k N

•

Jun 4, 2019

Going from the Basics of Logistic regression-Neural network -regularization- hyperparameter selection and finally knowing how Tensor flow makes it all come together is just brilliant.

I feel confident as I understand the basics well before using a framework that makes it so easy to execute.

By Basel A

•

Aug 5, 2018

Realy recommended for those who finished the first course and/or the machine learning with Prof Andrew Ng. A good in deep exploration of different topics in regularization. An efficient introduction to TensorFlow which will put your feet on the first step of using DeepLearning platforms.

By Miaoyin W

•

Oct 2, 2017

Need some improvement! I think the course is a little bit rush, especially on the 3rd week. I really like the 'test' assignments, which helps me to clear out a lot of important concepts. But the programming assignments sometimes bothers me not in the way of programming, but in the way of

By Hari K

•

Oct 6, 2020

Excellent! Very practical intro to tuning and improving neural networks. The four star is because I felt the programming exercises could be harder and not so much fill in the blanks. I don't think I would be able to code a adam optimizer or momentum from scratch given an empty py file.

By Juan A O G

•

Sep 4, 2018

A great course indeed! I give 4 star only because I'd have liked more lectures and programming exercises about Tensorflow, and how to train models using GPUs. Similarly, it'd have been great if Andrew explained in detail how to implement Batch normalization using computational graphs.

By Joseph D

•

Jan 13, 2018

Great course. Thanks for making it available.

I would have enjoyed more tensorflow lectures to help understand the underlying mechanism of the platform. I suppose the intention is to provide that understanding through the assignment, but more discussion in the lecture would be nice.

By Nicolas M

•

Mar 20, 2018

Good course but it would be interesting to add some other methodologies on learning rate ("Cyclical Learning Rates for Training Neural Networks", "Snapshot ensembles") and some explanations on categorical variables and embeddings matrix ("Entity Embeddings of Categorical Variables")

By Eloi P

•

Sep 16, 2017

Great course giving insight on how to fine tune deep neural networks. I believe the contents need to be a bit polished but that's totally understandable given its early stage. The comments in the discussion group will for sure help to fix some typos and make this course even better.

By Tuan N M

•

Mar 6, 2021

This course helps me a lot in tuning hyperparameters in training machine learning model, just one issue is the last programming exercise when using framework the guide is missing something, which is hard for some to complete, for me, I have to use Google Search to find the solution

By Siddharth K

•

Jul 15, 2019

Need Information about other parameters like #of iterations, how to choose number of hidden layers?, number of neurons in hidden layers, inclusion of few other strategies to choose neural network model will be helpful. If they are covered in next courses, then please ignore.

Thanks

By Sothiro P

•

Aug 5, 2018

A useful class delving into the nuts and bolts of building a reliable nn. Well structured and explained. I feel like the use of Jupyter in the homework makes it simpler than it should be. A large portion of the code is already written and the instructions often give up the answer.

By Mathieu B

•

Jul 11, 2020

For a person, who know a little on deep learning, I learned lots of things or, at least, got a clearer view on many concepts. A little reproach on the notation system : question on quizz sometimes might not be very clear for me - and the flaws of the grader on the assignments.

By Shan P S

•

Dec 8, 2018

A very good course for taking understanding of Deep learning one level above the basics. The course is theoretical, but the team has done their best to make it as much hands-on as possible.

I did face some intermittent platform issues with saving and submitting my assignments.

By Ramprakash V

•

Aug 3, 2020

Helps to have a structured approach towards tuning the hyperparameters rather than randomly doing. Also the course also helps understanding why such tuning is necessary and what improvements are being made in the model. Useful course but not suitable for beginners in ML/DL.

By Vasilii D

•

Dec 23, 2019

Material is awesome like all courses professor Andrew does. But (a) programming assignments are in style 'fill a couple of lines in 90% ready code' instead of end-to-end developing with guidelines and (b) there are a lot of mistakes in subtitles, assignments and even videos

By Matt G

•

Apr 22, 2022

The theory was good, but I think jumping to tensorflow at the end wasn't a logical, progressive step forward. They should have solidified the concepts more thoroughly, rather than jumping to the Tensorflow API. One would really want to have Tensor flow in a separate MOOC.

By Varun K M

•

May 19, 2020

A lot of content was repeated from the Machine Learning course by Andrew Ng on Coursera. Also, more on TensorFlow and other frameworks implementation would be interesting to learn. But at the end of the day, I did learn a lot of interesting aspects of deep neural networks.

By Maciej B

•

Aug 22, 2017

Course is very good especially when revealing "secrets" of various optimization techniques. Once again programming excercise is rather easy to pass as you are guided step by step so there is no space for serious mistakes. More "open" excercises/chalenges would be desirable

By Ruchita R B

•

Jul 20, 2020

This one took a little longer than usual to complete, It took more willpower to come back to it and continue in the course. It seemed harder, or explained lesser than the first course. Nevertheless, after spending extra time on it, Ive finally completed it. Thanks Andrew!