Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization に戻る

4.9

星

43,181件の評価

•

4,645件のレビュー

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 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

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.

フィルター：

by Adail M R

•Sep 13, 2017

Once more, Prof. Ng show in his simple style how to tackle the tough subject of hyperparameter tuning, pointing to several techniques and helping us selecting the most appropriate ones for the task at hand. The Tensorflow introduction is also very effective and engaging! Looking forward to advance my knowledge and experience with the next courses!

by Diego A P B

•Mar 06, 2018

Hyperparameter tuning and the other techniques seen in this course are not perceived to be the most fashionable areas of machine learning and deep learning. Nonetheless, they are crucial parts, and thus the techniques shown in this course will show you how to save great amounts of time and headache when trying to improve and finetune your models.

by Nestor H

•Jun 06, 2018

It was a great course to take. I could grab basic knowledge on TensorFlow and on some optimization techniques. I consider all the optimization algorithms are based on gradient descent, it is just that they tweak some parameters, but they are gradient-descent like algorithms. In summary, Dr. Ng is a genius and it is worth taking all his classes.

by Jay P G

•Dec 31, 2019

After knowing the basics of Deep Learning and Neural Networks (From the course 1) , this course explains the crux of improving and tuning of the neural networks and it's parameters and Hyper parameters . And the intro to tensor flow at last was just awesome(not exaggerating it!!!) . Congrats to Andrew and his team for such an awesome course .

by Shivdas P

•Dec 24, 2019

This course extends what has been taught in the preceding course, especially the different hyper parameters and optimisation strategies. Getting started with TensorFlow in a complete end-to-end example has been one of the things I was looking for and this course puts all that and many other things into perspective. Thanks Andrew and team !!

by Tamas K

•Aug 03, 2019

The course was great, thank you! However, I'm really looking forward using Tensorflow in C++ or Swift. The obscure, untyped nature of Python facilitates cargo-cult habits, creates some mystic fog around the variables (since it's not explicit if e.g. 'cost' is a concrete float or an entire computation waiting to be executed) and error-prone.

by Eulier A G M

•Aug 31, 2019

The course is very well structured, most of the topics here is perhaps kind of boring due the lack of real-problems projects, but if you stick to it and learn the concepts, will boost your understanding when using Deep Neural Network Frameworks, such as Tensorflow. That makes creating DNN easy to set, understand and apply to your problems.

by Suhas P

•Sep 21, 2017

Introduction to TensorFlow was wonderful. This course has helped me visualize and experience end to end flow of an actual machine learning project that helped a lot. Thanks to Andrew for taking efforts to design the course in a user friendly way. Programming tips are intuitive, helps save your time and allows you to focus more on learning.

by Chandan N

•Nov 27, 2019

Great insights into the theory of regularization and famous optimization algorithms like RMSProp and Adam. Helps in developing intuition regarding these algorithms work and implementing them from scratch was pretty rewarding as well.

As usual, Prof Andrew Ng patiently explains the theory and helps in building understanding of the material.

by Saransh M

•Aug 20, 2019

Started from the basics but made sure that they provided an in depth understanding of some very important concepts like hyperparameters and regularization will well structured quizzes and interesting programming assignments. Really liked the course and would suggest it to anyone trying to set their feet in the field of ML or Deep Learning

by Shuvayan G D

•Jun 16, 2019

This is probably one of the best courses on hyperparameter tuning. Along with Andrew's teaching , the course assignments are just perfect to get the perfect intuition of how optimizers work in the deep learning frameworks , also you will be able to build your own optimizer from scratch after doing this course , though not recommended. : P

by MOHD F

•May 17, 2019

This is an amazing course, it helps me a lot to gain the basic intuition, and the idea behind tunning our model, this course provides understanding basic maths of how we can knob various hyperparameters, which would eventually lead us to a better statistical model in term of both speed and performance... Thankyou coursera ...Thanks Andrew

by Ivan T

•Feb 04, 2018

Another great course! Enjoyed it very much. Learned a lot of useful techniques. One thing that could be nice to do is to add references and optional material for students who want to go deeper (i.e. add references to publications related to some technique or a blog page). Looking forward to taking more courses in the Deep Learning series!

by Jonah N

•Jun 04, 2018

The course really gave me insight into some of the optimization methods that are commonly used. It also helped me to get a better understanding of Tensorflow. I think y'all have done a good job presenting the information with just the right amount of math and explanation. I have recommended this course series to multiple friends already.

by Arpit B

•Sep 11, 2017

Thanks Andrew, As always you have been a superb teacher, I am very happy with the content of the course.

One suggestion is to increase the level of difficulty in assignments. Or you can have one more course to develop an difficult deep learning application from scratch, through which we can all apply the concepts and tricks you explained.

by Yan

•Apr 13, 2019

Although the concepts of deep learning ( ie. the gradient descent, the chain rule ) are quite easy-understanding and clear to most people, how to choose the hyperparameter and how to effectively carry out the projects are real essence. That's what I learn from this course. Thanks for so many genius researchers contributing to this area.

by Jayant R

•Feb 24, 2019

I didn't knew much about different optimization algorithms and how they work. This course helped in understanding those concepts. Also leaened how to tune hyperparameters. Now, I am able to read tensorflow codes on net and also able to write basic code. Prof. Andrew Ng is the best. Concepts gets very clear on first time watching video.

by Edwin G

•Dec 12, 2017

Some of the coding at the end was pretty tricky and I had to use the forums for help. That's what they're there for of course but I don't think the introduction to Tensorflow syntax was really sufficient - or maybe there could be some more optional help or resource to look through to help. Still very interesting and rewarding course!!

by Virginia A

•Jan 07, 2020

highly illuminating. Finally, with this second course, I could grab the deep concepts and conseguences of many terms I heard so many times during talks between data scientists. I feel now I could easily use what learnt to participate actively to those meetings and practically try things out on my methods and make them perform better.

by Karan S

•Apr 27, 2019

I'd been working on Neural network Models in my undergrad projects, but really couldn't answer much of the problems that I faced. The title isn't too appealing, because no new Network Architectures are taught, but in my opinion, this course is on par with the previous course on building Deep Networks from scratch. Highly Recommended.

by Anand K M

•Feb 10, 2018

A very nice course providing intuitions and concepts for tuning the hyper parameters in a neural network.

Also, provides a taste of using Tensor Flow (Neural Network Framework) in a comprehensive manner.

I would give my deepest thanks to the instructor, Prof. Andrew Ng for his invaluable time for building the course for the learners.

by OMAL P B

•Feb 24, 2020

This is an wonderful course for people who want to learn about the ways to improve their models and make their best and more robust. Andrew Sir makes the math behind the scenes very easy to understand. The course is easy to follow as it gradually moves from the basics to more advanced topics, building gradually. Highly recommended.

by Tanveer M

•Aug 28, 2019

Professor Ng's very clearly put a lot of thought into breaking down deep learning into the most understandable way for students around the world and it shows through the quality of this course. I cannot recommend this course enough to anybody who is looking to do machine learning, or simply understand the process from a high level.

by Bernard O

•Oct 24, 2018

The tips and great guidelines one gets from this course are gems in their own right. Practitioners in particular will get to appreciate all the usable advice to improve their neural networks and at the same time get to understand the principles behind the scenes on what truly drives those optimizations. Highly recommended course.

by Sinkovics K

•Aug 15, 2018

The course material is detailed and comprehensive and presented in a very digestible way. I felt like the home works lack a few topics (e.g. learning rate decay implementation, which I would find to be a useful exercise), but they give you a good understanding of what is going on. All in all, I definitely recommend this course!