|Applied Machine Learning On Coursera|
|Written by Sue Gee|
|Monday, 26 August 2019|
Coursera has added another Machine Learning Specialization. Its distinguishing feature is that is targeted at those working in finance, medicine, engineering, business or other domains where machine learning is taking hold.
Disclosure: When you make a purchase having followed a link to Coursera from this article, we may earn an affiliate commission.
We've all heard the buzz around machine learning and the way it pervades every aspect of our lives, see Machine Learning For Cat Control, and even that it pays well, see Machine Learning Engineer Rated Best Job 2019. As far courses as are concerned we have looked at offerings from Udacity, edX and Google as well as Coursera.
Coursera already has several options for studying machine learning and it seems a particularly appropriate platform for learning about it given that was co-founded by Andrew Ng. Ng's seminal course Machine Learning is still going strong having gone through many state changes over the years. Originally delivered face-to-face at Stanford University, where it was one of the most popular courses offered by the Computer Science department, it initially reached a wider audience as a series of lectures on You Tube. Then in 2011 it was one of Stanford's Free Computer Science Courses which propelled the MOOC explosion. For this initial presentation Ng adopted a new online format in which lectures were delivered in convenient segments of around 10 minutes with multiple choice questions interspersed.
Although overshadowed by the numbers enrolled in Introduction to AI taught by Sebastian Thrun and Peter Norvig, which had 160,00 signups and issued over 23,000 certificates for successful completion, this online experiment was a resounding success with 104,000 enrollments and 13,000 certificates. It also led to the formation of Coursera and was among the first classes to be offered there.
Machine Learning has now had nearly 2.5 million enrollments and is still available as a free course, with its next starting date being September 2nd. It is an 11-week course but as it has flexible deadlines you can take longer if required. It has a rating of 4.9 out of 5 based on almost 112K ratings and I can personally vouch for its excellence. Programming assignments are in MATLAB or Octave, a free open source alternative.
Initially all Coursera courses were free and issued certificates. Then it introduced paid-for certificates. Then along came Specializations in which students complete and gain certificates in a series of short courses (typically 4-6 weeks in length) and then pull all their learning together in a Capstone project and earn its final Certification. Until fairly recently, if you didn't want certificates, courses comprising a specialization could be audited for free, although this often meant that while you could view the videos and other course materials you couldn't complete the quizzes and programming assignments. While much of the content of older Specializations is still available for free new ones require payment on a monthly subscription basis ($75 per month) that lets you access all the course in the specialization. You have the benefit of a 7-day free trial during which you can cancel your upfront payment and financial aid is available.
The new Machine Learning: Algorithms in the Real World comes from the Alberta Machine Intelligence Institute. Although billed as a Specialization, and leading to a Certificate, there is no Capstone project and each of the four, nominally 4-week courses, are short and involve watching very short video snippets, plus readings and quizzes per week. The whole specialization is expected to require only one month to complete, with around 19 hours effort. It has fairly minimal prerequisites of a background in analytics, math (linear algebra, matrix multiplication), statistics and beginner level python programming.
By completing all four of them, participants will:
Details are available of the first two courses.
There is no syllabus as yet for the other two courses. The third course Data for Machine Learning will highlight how data it is critical to the success of your applied machine learning model and draw attention to bias and feature engineering. The final course Optimizing Machine Learning Model Performance is intended to synthesize everything your have learned in the applied machine learning specialization. claiming:
By the end of this course you will have all the tools and understanding you need to confidently roll out a machine learning project and prepare to optimize it in your business context.
or email your comment to: email@example.com
|Last Updated ( Friday, 22 November 2019 )|