返回到 Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization

4.9

星

42,330 個評分

•

4,518 條評論

This course will teach you the "magic" of getting deep learning to work well. Rather than the deep learning process being a black box, you will understand what drives performance, and be able to more systematically get good results. You will also learn TensorFlow.
After 3 weeks, you will:
- Understand industry best-practices for building deep learning applications.
- Be able to effectively use the common neural network "tricks", including initialization, L2 and dropout regularization, Batch normalization, gradient checking,
- Be able to implement and apply a variety of optimization algorithms, such as mini-batch gradient descent, Momentum, RMSprop and Adam, and check for their convergence.
- Understand new best-practices for the deep learning era of how to set up train/dev/test sets and analyze bias/variance
- Be able to implement a neural network in TensorFlow.
This is the second course of the Deep Learning Specialization....

Oct 31, 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.

Oct 09, 2019

I really enjoyed this course. Many details are given here that are crucial to gain experience and tips on things that looks easy at first sight but are important for a faster ML project implementation

篩選依據：

創建者 Amit G

•Nov 06, 2017

there is a lot of materiel that is being discussed during the lectures, and all of it seems like it could be really relevant. I am missing a consolidated course deck - ie something like a deck of slides on all the important concepts that are being discussed, for reference.

創建者 Kiet L

•Aug 27, 2017

Another awesome course by Andrew. I wish he was my professor in my grad school. I hope Coursera publishes all the notebooks + data on public github so I can redo all the exercise again. Too much info to digest in short amount of time. I can't wait for RNN and CNN courses.

創建者 Syed M H J

•Jan 09, 2019

Easily the best course on diving under the hood of how a Neural Network actually works and how to tune to the satisfaction of our results.

A no brainer for sure. The best part the exercises. You MUST do the exercises to understand thoroughly how the systems actually work.

創建者 WAN L

•Aug 01, 2018

I like this course, it details basic while popular technique we need to optimize neural networks. also the lectures on different optimization algorithms are very helpful for you to know details on how they run when we choose these algorithm in frameworks like tensorflow.

創建者 Yash M B

•Oct 22, 2019

Quite detailed curriculum. It is a great continuation for course 1 of this specialization series. As usual, Prof. Andrew Ng is there to guide our way throughout the course duration. A really fun and intriguing course which can lead to course 3 as a proper continuation.

創建者 PeterStephenson

•Jun 26, 2019

This course was perfect for me. I thought it was a good balance between theory and practice. I don't think I'm ready to start building NN's from scratch, but at least now I know how to get started. Also, I now have an understanding of the complexity of a ML project.

創建者 SHUBHAM G

•Jun 18, 2018

Mini Batch/Adam Optimization concepts was very well explained. I was expecting the detailed derivation of the backpropagation for the batch normalization case. Overall it was a great course and it greatly improved my understanding about concepts used in deep learning.

創建者 Favio A C

•Nov 03, 2017

4.5/5 A diferencia del primer curso que es una continuacion del de Machine Learning de Andrew Ng , aqui vemos una evolución del contenido , se pasa a ver miniBatch Gradient Descent, Regularizacion , Momentum , Adam , y un inicio a tensorflow

realmente un MUY BUEN Curso

創建者 Huaishan Z

•Oct 01, 2017

Through the class, the tuning of Hyperparameter is detailed introduced and more important is that why it's tuned is very clear. Suggest persons study deep learning to study this class carefully.

Expect to have more info from the current study in University or College.

創建者 John R

•Jul 24, 2019

I guess the difficulty is what you make of it, with further studying and dedication, but I would like to encounter more challenging assignments, where one has to code entire cells for instance, as opposed to a single line here and there.

But everything else is great!

創建者 Janzaib M

•Mar 04, 2018

Contains very good understanding of Hyperparameters and their tuning process.

Secondly, teaches very well the mathematics of optimizers such as GD, SGD, GD with Momentum, GD with RMSProp and ADAM.

Finally, a small glimpse of Batch Normalization.

Highly Recommended!!!!!!

創建者 Frank I

•Aug 25, 2017

I had previously used optimizers with momentum and variance momentum (Adam) with the understanding that they helped without knowing exactly how. This course cleared up all those tiny details and has left with with a greater appreciation of neural networks in general.

創建者 Thomas N

•Oct 09, 2019

This course broadened my understanding of what really happens when driving the cost function closer to its minimum and techniques to go there faster. I found this course instructive and the programming excercises helped a lot to digest the learnings from the videos.

創建者 Saikiran K

•Aug 03, 2018

I know deep learning already, but I saw many people who even know it doing this specialization,so i too started like that..but its a very good experience concepts are very well explaining and I am enjoying assignments a lot it a very fun experience doing all again..

創建者 Narek A

•Oct 08, 2017

I find this course very useful, many complex ideas are presented in a very understandable way! This course is like a collection of all important aspects! However, homework could be more difficult, because now almost all the answers are given in the python notebooks.

創建者 Sagren P

•Sep 04, 2017

This specialisation is an exciting journey - can't wait to start the next course. The foundational concepts of neural networks are expertly packaged in these courses, together with enough practical exposure to get you started on a fun learning and career experience.

創建者 Neil S

•Jun 17, 2019

Wonderful course that teaches one the intricacies of training better models. It's also great when learning to implement a neural network through Tensor Flow for the last assignment and realizing that you have a good understanding of whats going on "under the hood".

創建者 Michael S

•Aug 05, 2018

Overall, this is an excellent course, although it is not perfect. Trying to understand what is wrong when full credit is not earned for quizzes or programming assignments is sometimes "challenging". It would sometimes be useful to have more informative feedback.

創建者 Ertu S

•May 18, 2018

Great course., excellent well to the point, Only nuisance I observed is during submitting coding assingments required multiple tries since at first time, all the code somehow does not go thru. So needed to save and restart notebook and cut& pasted again. Thank you

創建者 Белоусов А Ю

•Sep 23, 2017

Great course. I really like it as it get more and more practical.

Few things might be missing from the class - it might be worth to encourage students play with algorithms a bit more. Say get back to the previous stage and add regularization to get better results.

創建者 Christos Z

•Apr 30, 2018

Grate course, only criticism is that week 3 didn't thoroughly explain how batch normalization parameters (gamma and beta) get updated during gradient descent. (i.e. how to get dgamma and dbeta). It could have been an optional lecture for the mathematically savvy)

創建者 Abdelrahman A

•May 19, 2019

it is wonderful course i learned more in Deep learning and how to apply regularization

and how to optimize cost function also programming in Tensor flow

i thanks all teaching assistant for there efforts to learn us

and i recommend this course to DL beginners

創建者 manish m

•Apr 28, 2018

I recommend everyone to go through this course if you really want to learn detail about hyperparameter tuning , optimizers and regularization used to make neural network better. It helps to open black box of Neural network and know in detail about how all works.

創建者 Lee

•Sep 05, 2017

Some very useful insights into practical implementation and optimization of neural networks, and a very welcome introduction to TensorFlow. After coding networks in numpy you both appreciate the framework, as well as understand what it's doing behind the scenes.

創建者 Sebastian E G

•Aug 18, 2017

Again, fantastic. Great way to explain how to tune your algorithms to improve bias and variance. Great explanation of what optimizers are used and how they function. Glad to know the nuts and bolts of the parameters usually defined in machine learning frameworks