Chevron Left
Back to Machine Learning Data Lifecycle in Production

Learner Reviews & Feedback for Machine Learning Data Lifecycle in Production by DeepLearning.AI

4.3
stars
808 ratings

About the Course

In the second course of Machine Learning Engineering for Production Specialization, you will build data pipelines by gathering, cleaning, and validating datasets and assessing data quality; implement feature engineering, transformation, and selection with TensorFlow Extended and get the most predictive power out of your data; and establish the data lifecycle by leveraging data lineage and provenance metadata tools and follow data evolution with enterprise data schemas. Understanding machine learning and deep learning concepts is essential, but if you’re looking to build an effective AI career, you need production engineering capabilities as well. Machine learning engineering for production combines the foundational concepts of machine learning with the functional expertise of modern software development and engineering roles to help you develop production-ready skills. Week 1: Collecting, Labeling, and Validating data Week 2: Feature Engineering, Transformation, and Selection Week 3: Data Journey and Data Storage Week 4: Advanced Data Labeling Methods, Data Augmentation, and Preprocessing Different Data Types...

Top reviews

SC

Jul 2, 2021

Interesting material. There are quite a lot of typos and many code snippets are directly from the tfx manual pages however the instructions provided and logic of the course is clear.

DD

Jul 20, 2023

Liked it for the most part. It was a bit dull when going over the details of schema updates and meta data. But that might be the nature of the beast.

Filter by:

26 - 50 of 166 Reviews for Machine Learning Data Lifecycle in Production

By Albeiro E

•

Aug 1, 2021

Thank you so much to DeepLearning.AI. You inspire me! This course is a key step that most part of enterprises should follow in order to construct robust ML systems

By ALAN S S

•

Nov 9, 2021

Incredibly useful and well teached. Awesome hands-on guide delivered by Robert Crowe, he is indeed a master. Im looking foward to learn with him again!

By Adarsh W

•

Oct 14, 2021

It is a very informative course. I learned a lot about data, metadata, schema and feature engineering, Also, Robert Crowe sir is a very good teacher.

By Flurin G

•

Sep 8, 2021

Lessons are well structured and clear, and the labs are very instructive. Above all the course is fun!

By Thiago P

•

Feb 1, 2022

Really nice course, but too much focused on only one framework

By Enrique C

•

Jan 4, 2022

Good intro but it looks like in other courses from deeplearning.ai, while they teach you something, they also try to "sell" people a specific framework. In this case, they seem to be selling TFX, whose API seems to be in constant flux with no guarantee (maybe not effort at all) of backwards compatibilty. It is very likely that if you download a notebook and try it in your computer, unless you're using the same library versions, it would not work. Some quizes seem to be not in sync with the lesons content (questions are about the content off the next session). not acceptable for a platform like Coursera that has horrible customer support and that is ruthless with users that have issues with their payment method.

I still recall how they sold people the Trax library in the NLP specialization which seems to have replaced Trax with huggingface. I take what is useful from these courses but I distrust their agenda.

By Antonis S

•

Mar 9, 2022

+ New cool way of working with many possibilities

-Many new concepts and code with no clear connection to the "known" way of working.

-New code concepts not very clearly explained Urgent suggestions for improvement: Make the new concepts and code clear to the audience. Connect the examples to the previous way of ML

By reto w

•

Nov 10, 2021

I was not happy with the course. In the part 1 the lecturer showed a lot of real world example of developing big ML-systems. The lecturer for this course is more a library creator than a user of it. And therefore also it feels like an advertisement for tensorflow. Which is an odd combination for me. So it does not teach a lot of useful theory because it focuses on how tensorflow manages pipelines and not a lot about the concepts. But also the programming examples are very artificial examples taken from the tensorflow tutorials or documentation. What I liked in the first course was the practical view on a specific problem. The programming exercises I also did no like because I did not learn anything useful. I only "learned" to use tensorflow a bit. But the concepts implemented are so basic that they are not interesting at all. I am aware that this has to be like this if we are not expected to program for two day but I don't see the benefit for me of solving mandatory useless exercises. The result of this was: I was skipping through the videos in 2x and was solving the quizzes as fast as I could. Speaking of quizzes. There were quizzes asking questions never mentioned in the videos and once the quiz was posed before the video where the things were explained. Also the quizzes used unusual wording for concepts plus not clearly written questions. In the end there were some useful insights here and there but it was quite an effort for me to filter them out as my motivation was lacking after some time.

By Tman

•

Apr 4, 2023

Well, I am a big fan of Andrew Ng, his initial ML course is what kickstarted my career change from a computer scientist to an established data scientist, I quite liked the Deep Learning Specalization, but this course is absolutely not what I hoped it would be. A whole week of explaining what feature engineering is? A data scientist would already know. Generally, there is very little MLOps-related content, and a lot of generic data science knowhow. The whole course is very heavy on tfx, which seems like a horrible tool to me. And finally, it is rather ironic when the notebooks spew one warning after another, and the notebook authors keep telling you to 'just ignore the warnings'. Not a thing that should happen in an MLOPs course (even if it is a minor detail).

By RN

•

Dec 24, 2022

It is clear that this course is focused on the use of Google's TensorFlow Extended (TFX) and Google Cloud, Kubeflow, and Apache Beam for the production of machine learning models. While this may be useful for those looking to work specifically with these tools, it may not be applicable to those seeking a more general understanding of MLOps skills.

Additionally, it seems that the primary purpose of this course may be to market Google's products, rather than providing a comprehensive overview of the machine learning data lifecycle in production. As a result, those seeking a more general understanding of MLOps may want to consider looking for a course that covers a wider range of tools and approaches.

By Simon B

•

Sep 8, 2022

By far the worst course from Deeplearning.AI. The slides and lectures are not condensed (as they usually are with Andrew Ng) and the lecturer repeats himself over and over. The slides contain so little information that one is actually better of not looking at them at all. Disappointing

By Meiting L

•

May 28, 2023

To be honest, at first I was a little bit disappointed watching the videos. Unlike Professor Wu's teaching materials, Robert's videos focus a lot on "what" and "why" (which seems pretty obvious most of the time), instead of telling us "how" exactly we should do to perform certain tasks. BUT, this missing piece is completed after the ungraded labs are finished. The ungraded labs provide a very comprehensive elaboration on what/how should certain task be done, with practical off-the-shelf functions. Please make sure you read the ungraded lab to make the best use of this course. It's really rewarding.

By Pablo S

•

Jul 26, 2023

Wonderful course, with great labs that really give understanding of the processes involved. TensorFlow Extended(TFX) functionalities are really state of the art. Although TFX manuals are not that great for the first time users, the labs help a lot in understanding the functionalities and uses.

By Guillem B P

•

Nov 6, 2023

Thorough exposition of the principles and concepts behind managing the data in a production machin learning model. I especially appreciated the detailed walkthroughs on the TensorFlow tools that manage the schemas, metadata and examples.

By Jakub Z

•

Nov 30, 2023

Very thorough and Specialized course for MLOps. Even though one knows and sees the big picture of the addressed issues, it results to be a wide range of topics which one doesn't know

By Erik H

•

Nov 5, 2022

After a few hiccups it was actually very good, because you proper had to dive into the documentation of tax to find out the solutions, which is usually the best way to learn it.

By DeeAnn D

•

Jul 20, 2023

Liked it for the most part. It was a bit dull when going over the details of schema updates and meta data. But that might be the nature of the beast.

By Shreyas R C

•

Jul 21, 2021

Best course for the professionals looking to upgrade there ML skills at production level! Thanks to the brilliant and wonderful course instructor.

By ANMOL K

•

Feb 1, 2024

very helpful and best for beginners. My only suggestion to beginners is to take the assignment and ungraded lab seriously.

By Youngjeon L (

•

Sep 11, 2021

Nice, Awesome MLOps Pipeline with TFX! I recommend this course anyone who want to build ml pipeline! Good Luck! :)

By Nam H T

•

Jan 16, 2022

Great course with useful exercises to get learner familiar with ML Data pipeline using TensorFlow Extended!

By simegnew A

•

May 26, 2023

It is an excellent course if you have a machine learning and computer vision/image processing background.

By Fernandes M R

•

Jun 19, 2021

Its good, I think was a little difficult because TensorFlow, but it was very explicative.

By Luis S S

•

Sep 10, 2021

Excellent course. Theory and practice well combined, to fit diverse curiositiy levels.

By Han B

•

Jan 15, 2022

instruction on debugging jupyter and submission issue is important for learners