|TensorFlow For Beginners From Coursera|
|Written by Sue Gee|
|Friday, 08 March 2019|
The first course in a new Machine Learning Specialization from has just made its debut on the Coursera Platform. An introduction to TensorFlow, the course is a collaboration between Andrew Ng's company, deeplearning.ai, and Google's TensorFlow team.
TensorFlow is an AI framework which came originally from Google and is now an open source project on GitHub. Enthusiastically adopted for Machine Learning, it is widely used and already has lots of educational resources.
Introducing the new Specialization, TensorFlow: From Basics to Mastery, Andrew Ng writes:
If you want to build scalable AI-powered algorithms, you need to understand how to use the tools to build them. This course will teach you best practices for using TensorFlow, a popular open-source framework for machine learning.
Introduction to TensorFlow for Artificial Intelligence, Machine Learning, and Deep Learning is a self-paced course that is nominally 4 weeks in length with 4-5 hours per week. It is at Beginner level although you do need experience in Python coding and high school-level math. Prior machine learning or deep learning knowledge is helpful but not required.
Like other courses in the Specialization this introductory module is taught by Laurence Moroney who is a developer advocate from Google Brain and uses the Google Colaboratory for the hands-on exercises.
Each week opens with a videoed conversation between Ng and Moroney. There's a surprise for Ng in the first one when Moroney reveals that he got into Machine Learning by taking Ng's courses and specializations!
Week 1: A New Programming Paradigm is intended to provide:
a soft introduction to what Machine Learning and Deep Learning are, and how they offer you a new programming paradigm, giving you a new set of tools to open previously unexplored scenarios.
After 15 minutes of video and a couple of short readings you get started with the Google Colaboratory in a short video and then tackle a short exercise using it, which is ungraded but available even if you only audit the course for free. If you want a certificate and pay for the course you'll also have access to the graded quiz.
In the other three weeks you use TensorFlow to tackle ever more sophisticated problems. Week 2: Introduction to Computer Vision is the starting point where you solve problems with just a few lines of code. Week 3: Enhancing Vision with Convolutional Neural Networks builds on this and Week 4: Using Real-world Images makes things more tricky by using complex images.
The course outro is a conversation between Moroney and Ng that reviews what has been achieved so far and encourages you to move on to the next course, where:
you'll go a little deeper into Convolutions, learning how you can go into depth with real-world images, and learn many of the techniques used in challenges such as those run by Kaggle!
At the moment I cannot find out what happens beyond the second course, other than that remaining courses will be released over the next several months - suggesting there are quite a few of them. Having dipped into this introductory course, however, it seems an attractive way for newcomers to Machine Learning who already know Python to get a feel for Tensorflow and see it in action.
The addition of a beginner-level Specialization means that deeplearning.ai has Artificial Intelligence/Machine Learning offerings at many levels on Coursera. For non-programmers, there's the recently launched AI For Everyone, which is aimed at managers and those who want to understand how AI can be incorporated into their organizations without doing it for themselves. For those at Intermediate to Advanced Level the Deep Learning Specialization comprises five courses, between 2 and 4 weeks each (77 hours in total). You can of course progress up the ladder as you learn more and the step between the new TensorFlow specialization and the the more advanced one is still Andrew Ng's ever-popular Machine Learning course, which, as I've said before, is highly recommended.
or email your comment to: email@example.com
|Last Updated ( Friday, 08 March 2019 )|