Вернуться к Introduction to Deep Learning

Отзывы учащихся о курсе Introduction to Deep Learning от партнера Национальный исследовательский университет "Высшая школа экономики"

4.6
звезд
Оценки: 1,296
Рецензии: 290

О курсе

The goal of this course is to give learners basic understanding of modern neural networks and their applications in computer vision and natural language understanding. The course starts with a recap of linear models and discussion of stochastic optimization methods that are crucial for training deep neural networks. Learners will study all popular building blocks of neural networks including fully connected layers, convolutional and recurrent layers. Learners will use these building blocks to define complex modern architectures in TensorFlow and Keras frameworks. In the course project learner will implement deep neural network for the task of image captioning which solves the problem of giving a text description for an input image. The prerequisites for this course are: 1) Basic knowledge of Python. 2) Basic linear algebra and probability. Please note that this is an advanced course and we assume basic knowledge of machine learning. You should understand: 1) Linear regression: mean squared error, analytical solution. 2) Logistic regression: model, cross-entropy loss, class probability estimation. 3) Gradient descent for linear models. Derivatives of MSE and cross-entropy loss functions. 4) The problem of overfitting. 5) Regularization for linear models. Do you have technical problems? Write to us: coursera@hse.ru...

Лучшие рецензии

DK

Sep 20, 2019

one of the excellent courses in deep learning. As stated its advanced and enjoyed a lot in solving the assignments. looking forward for more such courses especially in Natural language processing

AK

Jun 02, 2019

one of the best courses I have attended. clear explanation, clear examples, amazing quizzes & Programming Assignment this course is advanced level, don't enroll it if you are a new starter.

Фильтр по:

251–275 из 289 отзывов о курсе Introduction to Deep Learning

автор: Robert K

May 21, 2018

I've dived into this course only AFTER completing Andrew Ng's specialization "deep learning". In that sense this was a nice "revision" with additional set of exercises. Some of the topics introduced were nice exercise in ultimately "testing" your knowledge from other sources. Having said this, you really need previous exposure to machine learning, and I'd also say - deep learning.

But it doesn't give much beyond this point. Lecturers vary in terms of knowledge, or rather the ability to clearly present it. Coursera serves might not be enough for most exercises, and it pushes you to set-up your own machine (if you have a proper one) or configure one on the cloud. With many services it is rather easy now.

Overall, I recommend it as a review, an introduction AFTER some exposure. Some additional material might be new to you, but no necessarily if you followed other courses. I am more eager to look into further courses in the specialization.

автор: Ramin A

Jan 21, 2019

Overall I enjoyed the course, but it lacks structure. Some materials are assumed to be well known by the learner and surprisingly some easier ones are not. I like to see the math, but it needs more materials to support it. Most instructor's have very heavy accent and tend to speak too quickly, I find myself rewinding multiple times just to figure out what was being said. Homework's are not too difficult, and are enjoyable. Except for the last one where you need to wait for a peer review. I think this can be a flagship course with more efforts.

автор: Hermon A

Aug 12, 2019

The explanation of TensorFlow is not enough and the programming homeworks have already a lot of already written (because, i would be very difficult to programming the all of the homework by ourselves in this stage of learning). I think it is better programming homeworks with examples more easy, but with more programming by ourselves.

At least, I think it is already well enough for the final evaluation, the automatic correction and then, the correction by peer only delay the evaluation.

автор: RJ C

Jun 26, 2018

I could not understand what the lecturer in the second week was saying. Overall good content but awful presentation. Exercises are ridiculous, my code is working fine, but since I do not use the same function as teachers and I do not get the same result to 0.00001, I cannot pass the class. Definitely will not be renewing this class. Think twice before signing up..I am sure the guys that made the class are really smart, and the content is high quality, but overall I am disappointed.

автор: Juan C E

Feb 27, 2018

The quality of some of the video session is not good, especially for RNN's. Very general, badly explained and little practical information for the practical assignments. Yor have to "learn" the material, not just look for additional information, from other sources.

The pratical assignments are note always well designed, and some are full of flaws. After many many hours of dealing with some of them, you get the impression that you've passed the assignment but not learned much.

автор: Carlos V

Oct 07, 2018

The Course is good, probably should be called introduction to advance deep learning, the complexity of the assignments make you put lots of efforts around them, that is rewarding at the end, make sure you have plenty of time to dedicate to this Course, one thing the Course could improve on is to try to minimize the switch between libraries and the low-level coding with high-level coding between TF and Keras sometimes it creates confusion.

автор: Zhaoqing X

Jul 20, 2018

Well, I think it's a good course for introducing us to Deep Learning and it has better(tougher) assignments than Andrew's. It also covers more knowledge than Andrew's. But the quality of the course is not that good. The Russian accent is not important because my native language is not English as well, but the assignments are frustrating. The mentors cannot answer the questions that widely appears in the course.

автор: Jae L

Mar 01, 2018

Lecture slides need more written explanations, information, and math. Also, jupyter notebooks seriously lack information describing codes, explanation in neural network functionalities, and architectures. Please, practice clearer speech speaking, if it's hard to change, supply detail written notes to read for students.

автор: Marian L

Apr 12, 2019

I'm not sure that this course is needed at all. Folks are trying to explain multiple architectures of Neural Networks, without giving an actual understanding why it works. Plus I have a feeling that all of this things are going to explained in next courses of this specialization.

автор: Suewoon R

Nov 21, 2017

Great course and materials. I'm glad that I'm learning a lot from this course. I don't bother with different English accents but a couple of lecturers having too many pauses really keeps me from focusing on the lecture.

автор: Paco J A

Jan 01, 2018

Esta muy bien el curso en términos generales, pero para mi gusto falta explicación de algunos temas más en profundidad. Además, los videos de una semana en concreto cuesta seguirlos por el marcado acento hablando inglés

автор: Samson D

Jan 16, 2020

The video content is quite good and I've learned a lot. However, the preparation for the exercises is insufficient and the format of fill in the blanks is not really as educative as it is confusing.

автор: mohamed e a

Jan 21, 2018

there is some lecturers are not talking clearly stopping too much while explaining, the materials is very good with high quality

автор: Dizhao J

Mar 24, 2018

the pronunciation of the lecturer is not clear and too fast, hope they could speak slower and clearer

автор: Shubhra S

May 28, 2019

The introductory course is good. But I think you should include more reading material for RNN

автор: Kamil M

Apr 25, 2018

Neural network introduction is conducted very chaotic. Not all topics are explained well.

автор: Siwei Y

Dec 27, 2017

автор: Mathieu D

Apr 15, 2018

not enough content in the video to pass the exam.

автор: Kushagra P

Apr 12, 2018

The RNN week was very bad.

автор: 唐志强

Jul 30, 2018

автор: Shahzaib M

Aug 19, 2019

theory was good but at the time of Assignment i really felt blank as i have studied nothing, which i mean there is no technical support given in lectures, may be this is my fault that i cannot cop-up to the complexity. but still there is a room for improvement may be 5 to 10 min video to help student understand what they are supposed to do.

starting with coder decoder i literally gave up on assignment. so i had to search web and i felt lack of external matters too from where i could get help i am hoping that this response will the up coming student.

i focused on the things need to be improved but it does not mean that the course was not good over all. starting week was quite good i rate those week 5/5 stars. but later on the journey i had problem in understanding the pronunciation but than i realized its not that it is the material i am not clear of.

Thanks.

автор: David P

Dec 08, 2019

This is an extremely poorly prepared course in which the lecturers just throw material at you without bothering to make it even slightly comprehensible. One has to struggle really hard to understand what they are talking about -- they often use concepts and terms that have never been defined before, the slides are sloppy and often formulas make no sense. That said, I'll probably continue struggling with the lectures, since this is (unfortunately) the only advanced resource for deep learning at the moment. I really hope the lecturers will listen to the multitude of negative reviews and make an effort to improve their presentation.

автор: HAZEM D

Jul 04, 2019

good course with great lectures , but the assignments are very painful to complete , they are not hard but the training of the model takes to much time and the coursera notebook always crushes , it took me 1 week to finish an assignment after several trials of training the modal , i ended up by using google Colab with accelerated GPU in order to finish the assignments, also the instructions in the assignments are often not very clear. i suggest to reformulate the assignments and delete or modify the part where you have to train the model and wait several hours to submit the notebook .

автор: Stefano C

Jan 06, 2018

The course has a high potential, with large content and expert instructors. However, you soon realise that the content is indeed too large, and it's never covered with enough details and examples. The most disappointing part, however, is the assignments: the topics covered are too advanced and, as a matter of fact, the learner is only given the chance to implement a tiny part of it.

автор: Caio A A O

Nov 27, 2017

It started great, but became a bit too shallow. There's also little to no support from instructors, even when there are bugs.