Learner Reviews & Feedback for Structuring Machine Learning Projects by DeepLearning.AI

I love this course very much and I would strongly recommend this course to other DL colleague. It is truly that DL is a highly empirical process which needed to be more systematic. In this course, I have learned many methods to make DL more controllable and predictable. Nice Job! Thanks!

Well, once again I'm so happy and very satisfied of the content proposed in this Course by Prof. Andrew Ng. Thanks a lot for this valuable content and for always making it easy to understand. I'm getting more and more knowledge in DL which happens to be very usefull for my Phd project.

In my experience with Machine learning, we usually spend more time checking the algorithm than checking the best distribution of our data. In this course, Professor Andrew teaches us the need and obligation to create a correct distribution of our data with examples from the real world.

Extremely helpful for a beginner so as to think like a machine learning problem solver. I think there should be more quiz added to this course with scenario like given in two quiz. I have never enjoyed any course so much by taking same quiz again and again to get better understanding.

In the bloom of Deep Learning/Machine Learning industries, know how to build a project is more important and a priority to know what knowledges to build that project. Break the problems, take each step follow the guide and avoid common pitfalls in process, this course will satisfy you.

this is definitely the best course i had taken. it has just 2 weeks, but it was the hardest. i will definitely come back to see the teachings here explained to check up if i'm thinking correctly so i don't make much mistakes in taking a direction in projects.

definitely recommended!!!

The case study format for quizzes was highly effective in helping me uncovering gaps in my knowledge that I didn't know were there. I would have liked to see at least one more case study per week. One per week just wasn't enough.

Overall good course with a few minor video glitches.

I found this course really interesting because it gives many details on what path to follow to achieve better results not only depending on the amount of data we have but also taking into account some small details that can make a difference when starting machine learning projects.

A very well systematic approach explained, to structure ML projects.Can be grasped and implemented by anyone, let it be a beginner or some expert.Really liked the idea of case study in quiz. (Wait ! How about extending this idea into some coding exercise ? Would be some real fun !)

There is no coding in this course, but you learn a lot of how to design a Deep Learning Study. I learned a lot about the distribution of Training/Dev/Test sets and how to diagnose problems when a neural network is not performing as well as anticipated or if it is performing well.

Outstanding course. Many of the points made in this course mirror the hard earned knowledge I gained back when I worked on Dynamic Rank search engine focused neural networks.

This may end up being my favourite of the 5 courses but let's see if the last two have more math first. :)

This part of the specialization is short but it includes a lot of valuable information. Many of the tips are quite basic engineering best practices which most engineers should find natural, but some are very specific to deep learning and these are particularly useful to newcomers.

This is truly some information you'll never get in a standard class setting; this is more similar to compiling years of ML experience into short packets of advice that will guide your decisions for years to come. Extremely helpful, and recommended for all deep learning engineers.

This course is really what any PhD would need to conduct his research in more time saving and efficient manner. It would be great if coding was accompanied (even if only running and watching results) to touch all aspects of analysis and suggested improvements could be visualized.

Great course. I had a couple of ML courses at University, but nobody ever explained those concepts: orthogonalization, the data mismatch problem and what to do with it, different versions of human-level performance, end-to-end learning pros cons (everyone just talks about pros).

To Andrew and team (mentors and organizers), I am glad I opted for this course! You guys give such great insight on approaching and solving a Deep Learning problem, I don't think I would have ever found a better introductory course on Neural Nets. Thank you so much, everyone!!!!

One of the most important course in this series . This course actually helps you visualize the problems and standstills you might face when you are working on a model in real life. It also talks about practical solutions to improve your model that are valuable in the tech world.

An unique course. I don't think such a course is offered by any MOOC. I would suggest every DL enthusiast to take this course.

The programming assignments are very challenging. It forces us to think abstractly to find solutions encountered during real life Deep Learning problems.

Despite this course is labeled as basic level, it contiains very useful information related to strategy of developing ML projects. And use cases prepared by prof. Ng and his team is what you will get only by practice. It really helpful to structure what was learned by this day.

A great insight into how to improve the performance of the deep learning system without having to actually spend long hours/days and working on real project. Learnt a lot in improving the model's performance and where to look for the errors and how to invest time in debugging.

I totally agree with your flight simulator analogy. This really helps me to learn your experience in practising machine learning knowledge; which otherwise I need to spend many years of doing "try and error"

Thank you very much for your kind sharing of your practical experience

As always, very well structured material considering the nature of the content and trying to make it understandable and make sense. I also appreciate that it is rooted in real-life experience which serves to make me pay really close attention to everything that is being said.

This course teaches the trade off between Bias, Variance, Data Mismatch . You will learn how to split data and how to evaluate your model. It also covers error analysis systematically. It gives many examples of transfer learning, multi-task learning, and end-to-end learning.

Very refreshing to learn about also the more practical aspects of machine learning project like organising, structuring and executing the projects. The course definitely gives me more ideas now on what to do when starting a project and what to look into when facing problems.

Very useful guidelines for approaching projects! This topic is rarely addressed in comparison to the discussion of modeling techniques, however, in the real-world application, the trade-offs on where to start and how to proceed are just as important as the model themselves.