This course takes a real-world approach to the ML Workflow through a case study. An ML team faces several ML business requirements and use cases. The team must understand the tools required for data management and governance and consider the best approach for data preprocessing.
The team is presented with three options to build ML models for two use cases. The course explains why they would use AutoML, BigQuery ML, or custom training to achieve their objectives.
SG
A lot of core neural network topics were presented in a productive way and I particularly liked the LAB showing how to write custom estimators.
CK
Very interesting course! Specifically the last modules with custom estimators and the ability to create estimators straight from Keras!
RG
Very very informative for learning....had good time to gothrough this course...
RL
Sufficient theory to understand the basis of the ML approach with practical insights to help get started with building models
MH
some labs with bucket gave errors since it needs to specify some stuff from the bucket like region but no instruction about it
VK
Nice course however I think it suits folks who have good exposure of ML to take complete advantage of the techniques
MM
Iam very satisface, the contents, methodology, tools were exceletents, and the translations were very good im not was problems with the platform is very easy and intuitive
PT
It would be nice to have more complex datasets where predictions would be more meaningful.
KR
This course tells about the different aspect of usage of art and science in machine learning.
MJ
Good material on the art and science of ML - specially the custom estimator stuff.
PT
Great inside about the core topics with real word problems which make learning easier and practical. Thanks a lot for this great course
MB
thanks for the great work. There is so much to learn and I appreciate the effort you made to break things down and providing lab while making the hard decisions on what to commit.
显示：20/135
It has a lot fo great content, but the way it's laid out confused me greatly.
It also felt very rushed at times, many parts weren't explained in detail, making many parts much less educational than I feel they should've been.
The best parts are where Lak is presenting, even in the lab solutions he went to more details than anyone else and made even those videos interesting after solving it myself.
The quality of the lesson material is great but the quantity is nowhere sufficient to get the hands-on experience
Labs vms are to slow. Speaker is difficult to understand. Mic varies and speech pattern is not clear. The presentations need some graphics rather than a guy talking. Sketch out the ideas on a white board rather than talking 5 minutes to a single slide.
Labs were had less quality than in previous weeks.
My feeling is that this entire specialization is a glorified demonstration of what GCP can do with ML. The labs are not interactive and in some cases did not work. I don't feel that I have learned anything new. If I were to use GCP for ML purposes, I would need additional training to do it. I don't recommend this specialization
The best course ever to give you the glimpse of the whole ML land - what is ML? Why ML? and most importantly How to do ML for real life business problems - from developing models to model evaluation, serving in production with all scaling tips and techniques with essence of Cloud. The best course for ML practitioners out there. Learnt a lot and what other place to learn from apart from the people who are actually doing it for the most popular products used across the web and mobile world.
This final course is also very good. Embedding is my favorite part and Lak is my favorite instructor.
Thanks Googlers! Looking forward to the next five courses.
Too shallow to truly be useful. I think if anything it gives you an idea of what's possible and roughly the areas you should explore and learn about but you won't learn too much following this through.
Good course, but I couldn't get over the Estimator API. IMHO it's too complicated compared to Keras and I just could not force myself to care about it.
Some lab contents are distracting from the core subject, deviate from the video contents.
A good course but the one of the lab is not working. There's no way to skip it. Had to contact support for a refund. Hopefully it has been resolved.
i cant open a lab in the module number 4 and it give an access denied so i cannot take the certification and i have finished all the other module
There are several out of date labs that don't work and hence it is impossible to complete this course. It is a shame that after repeated attempts to get things corrected nothing seemed to work so I had to cancel the subscription and hence abandon the quest for the Google Machine Learning Engineer Professional Certificate.
Unable to complete Lab session. it shows as Sorry, access denied to this resource.
Waiting for 3 weeks for 1 lab to be fixed, yet nothing !
I have learned a lot of things about Machine Learning from this course. Thanks, Google Cloud team form this awesome course. Actually there are many Machine Learning courses or articles out there. But nobody tells me what is the best way to do one thing, why I will use this step for certain things. If anyone tells but he could not show the proper visualization, it is the correct way to do so. Now I will be participating in real-life projects to apply what I learn. If I get stuck then I will use this course as my reference guide.
This course goes deeper into how to improve your Tensorflow ML models' performance without going deep into the models themselves. A good intro to learn how to systematically tune hyperparameters like the learning rate or batch size, and also how to start using regularization techniques and embeddings. Finally the course also opens the door to custom estimators, keeping you interested for the next courses.
Good class with a lot of information and new terminology thrown at you. Especially if you don't have a machine learning background. The labs are done in a way where you can get familiar with TensorFlow and Python programming, without having to know Python programming. Good intro if you want to be familiar with GCP ML. If you're looking to write a bunch of code this will probably not be enough for you.
This is one of the best courses in this specilization, as it goes deeper in the internals of ML and Neural Networks. Also, it presents two important features such as Embedding (creating sparse tensors from features) and Custom Estimators, which is the first step to get into advanced topics such as implementing new estimators from Research Papers.
I've really leveraged from this course a lot.
Thank you!
The only concern is that There are lots mention of Batch Normalization however nor instructor neither lab exercises point out how to do proper Batch Normalization using Tensorflow. It seem like they suggest manual scaling. Also best to include a section how to use TensorBoard to fine tune training properly. Not clear how to monitor and correct training parameters thru TensorBoard.