All good things come to an end

The MOOC ended last week with Class 5. It will remain open until the end of Sunday, 20 Oct (all time zones). Statements of Completion will be emailed once the course closes. The course material will remain available up indefinitely at:

 

http://www.cs.waikato.ac.nz/ml/weka/mooc/dataminingwithweka/

under a Creative Commons Attribution 3.0 Unported (CC-BY 3.0) license. Use it however you like! We've also added the music, as MP3 files, to the course material.

We are planning to re-run Data Mining with Weka early next year (probably March). And your enthusiastic feedback has encouraged us to think about a more advanced version. But it does depend on support from Waikato University, and we are waiting until this course is over before commencing that conversation.

 

A survey about the course is available at:

 

https://weka.waikato.ac.nz/dataminingwithweka/assessment?name=Pos

 

Please fill it in! The response rate, and your feedback, as well as the course completion rate, will no doubt influence our ability to run future courses.

 

We have very much enjoyed giving this MOOC. And we have learned a lot in the process! We are preparing a report on the experience and will post a draft to WekaMOOC-announce, in case you're interested.

 

You can read more about Waikato's machine learning research group at:
 

http://www.cs.waikato.ac.nz/ml/weka/

 

Our publications are listed under the Publications tab. 

 

In the good old days, Weka was an externally funded research project. But that ended long ago. Both Weka and this MOOC are supported entirely by our Department and University. If you think these efforts are worthwhile and would like to support them financially, that would be lovely! Please do so here:

 

https://weka.waikato.ac.nz/dataminingwithweka/donation

 

All donations are directed to research: no administrative charges are incurred.

 

Finally, how about coming to Waikato to study? Our Department's web site at
 

http://www.cs.waikato.ac.nz/

 

has links to our research groups, and to graduate student information (MSc and PhD).

 

Excuse the advertising :-). Hope you had fun with the MOOC. Don't forget the survey. See you next time!

 

cheers

 

Ian

Class 5 is now available

The lessons for Class 5, the last in our course, are now available on the course website:

        https://weka.waikato.ac.nz/

Class 5 broadens out to consider some more general issues. It's a short week, with just four topics: 

• 5.1: The data mining process
• 5.2: Pitfalls and pratfalls
• 5.3: Data mining and ethics
• 5.4: Summary

 

The post-course assessment opens on Wed Oct 9. Everything will remain open for one further week, until Sunday Oct 20, when the course will be closed.

 

We will also ask you to complete a survey for your opinions of the MOOC. 

 

We will probably run the course again, perhaps in late summer -- I mean NZ summer! -- maybe in March 2014. And we are contemplating a follow-up MOOC that covers topics we couldn't fit into this course. But our ability to continue with MOOCs will depend on the success of this one, and the number of people who complete it successfully will be a key metric. Thus I once again urge you to do the assessments for my sake, if not your own :-)

Please keep up the help on the course forum -- we greatly value your assistance.
We'll soon put out a short (optional) video to discuss a few issues that have arisen in Class 4.

cheers, and enjoy the remainder of the course!

Ian

Class 4 is now available

The six lessons for Class 4 are now available on the course website:

 

https://weka.waikato.ac.nz/

 

This class introduces some more advanced methods and techniques. The topics are:

• 4.1: Classification boundaries
• 4.2: Linear regression
• 4.3: Classification by regression
• 4.4: Logistic regression
• 4.5: Support vector machines
• 4.6: Ensemble learning

The last three are high-performance contemporary algorithms. I aim to give you a conceptual understanding of what they do and how they work, but not the gory details. You have to learn to live and work in a world where you don't understand everything. You will see some mathematics in Lessons 4.2, 4.3 and 4.5. But don't worry: I'll explain it, and anyway you don't have to fully understand the math. 

 

Next week is the last. And it's short: Class 5 has only 4 lessons, not the usual 6. And it's more relaxed: no math at all.

 

It's great to see so many people helping each other on the course forum. Please keep it up -- we greatly value your assistance.

 

By the way, we've just published a blog post about the music that accompanies the course.

 

We'll soon put out a short (optional) video to discuss a few issues that have arisen in Class 3.

 

Ian

The music of the 'Data Mining with Weka' MOOC

Some people have been asking about the music that opens each lesson video of Data Mining with Weka. It’s the start of the Allegro movement of Mozart’s Divertimento No. 2, arranged for three clarinets by Michael Whewell.

Every week for many years now, clarinetists meet at my house to play music for fun. We joke that we’re slowly changing from a music group that eats beforehand to a dinner group that plays afterwards! We share food, wine, good company – and, of course, music. We play classical music, contemporary music, ethnic music, jazz – right now I’m obsessed with tangos by the Argentinian composer Astor Piazzolla. I have a large collection of music for clarinet duets, trios, quartets, quintets, and sextets.

Here’s the lot of us – six, along with some additional fans (including a garden gnome) – at Woodside House on beautiful Waiheke Island near Auckland, where the opening music was recorded.

 

 

 

 

The Mozart that opens Data Mining with Weka was played by three of us (I mistakenly imply four in one of the videos). Here we are, this time in my house. That’s Paul, me in the middle, and Sarah.

 

 

 

You can hear more music by going to my home page; I’ll play another tune every time you refresh the page.

 

Ian

 

 

 The music that closes each video, and the matching opening music for the Questions videos, is specially composed for Data Mining with Weka by Teresa Connors of DivaTproductions Inc.

Originally from the musically rich island of Newfoundland, Teresa studied both Composition and Opera singing at Dalhousie University (in Nova Scotia, Canada) as well as the Banff Center for the Arts. She has recently completed a Master of Music degree in the Music programme here at the University of Waikato.

More of Teresa's work is available on her Vimeo channel.


 

 

Class 3 is now available

The six lessons for Class 3 are now available on the course website:

        https://weka.waikato.ac.nz/

After this week there are 2 weeks to go (classes 4 and 5).

It's good to see so many people having fun with the course! A few errors in Class 2 Activity answers have been reported, and corrected in the MOOC.

The mid-course assessment is now available. Do it when you have finished Class 2 (although it will remain open for the rest of the course). The final assessment will appear during week 5.

My goal is to enable you to learn as much as possible from this course, and I recognize that doing the assessments may not be a priority for you. However, our ability to mount follow-up MOOCs will depend on the success of this one as perceived by my University -- and the number of people who complete it successfully will be a key metric. Thus I urge you to do the assessments for my sake, if not your own :-)

We will soon put out a short (optional) video to discuss a few interesting issues that have arisen in Class 2.

cheers, and keep going! Weeks 3 and 4 are the central part of this course.

Mid-course assessment is now open

The mid-course assessment is now open on the course website:

https://weka.waikato.ac.nz/

You can try the assessment as often as you like. It will remain open for the rest of the course.

cheers, and good luck!

Ian

Class 2 is now available

The six lessons for Class 2 are now available on the course website:

https://weka.waikato.ac.nz/

Following that, there are 3 weeks to go (classes 3, 4, 5).

It's good to see so many people having fun with the course! I hope that the teething problems with installing Weka, accessing the book excerpts, and locating the activities are now solved.

The activities are a crucial part of the course: they're where most people will do their actual learning! However, they do not form part of the assessment, so don't be scared to get wrong answers.

"Data Mining with Weka" has been designed so that participants at many different levels can learn as much as possible – and complete the course successfully. You don't have to do the reading. All you must do to succeed

are the mid-course and final assessments -- which you can try as often as you like. The mid-course assessment will become available this Friday (Sept 20) and remain open for the rest of the course. The final assessment will appear during week 5.

We plan to put out a short (optional) video later today to discuss a few questions that people have asked.

cheers, and keep going!
Ian

Welcome to 'Data Mining with Weka'

I'm pleased to announce that the course "Data Mining with Weka" has now commenced. The six lessons for Class 1 are available on the course website

 

 

https://weka.waikato.ac.nz/

 

 

We will release classes 2, 3, 4, and 5 at approximately the same time (Monday noon NZ time) in the upcoming weeks, and send reminder announcements. The course closes on Oct 20.

 

The course includes the following resources:

 

•  the Weka software; Lesson 1.2 gives downloading instructions
•  videos, one per lesson, on YouTube
•  the videos include captions, which can be turned on in YouTube
•  we recommend viewing in HD format, again a YouTube control
•  slides used in the videos (PDF format)
•  text files containing transcripts of the videos
•  activities that follow each lesson
• access to selected excerpts from Data Mining (3rd Edition) - plus you can buy a discounted copy from the publisher
•  mid-course assessment (opens Sep 20, at the end of Week 2) 
•  final assessment (opens Oct 9, during Week 5)
•  discussion forum
•  announcement forum, blog, twitter feed (available from the course website)

 

for Chinese participants:

•  videos on YouKu
• one version with captions in Chinese (another with English captions is available on our YouKu channel)

 

Some notes for participants:

•  work through the videos and activities at your own pace, in your own time
•  a new class appears every week; old classes will remain available until the course closes
•  in theory, you could leave all your learning to the last week (we don't recommend this!)
•  only the mid-course and final assessments count towards the Certificate of Completion
•  feel free to install Weka in advance, but please ensure that you have version 3.6.10
•  if you already know something about Weka, feel free to skip the first class (or two)
•  during the videos, it may help to follow with Weka on your own computer ("click along with Ian")
•  the course should take 2–3 hours/week (3–4 hours if you do the optional reading).

 

Please help us by filling out the pre-course survey if you have not already done so. Finally, note that running a MOOC is a new experience for us! -- please bear with us if problems arise.

cheers, and good luck

 

Ian

Enrolments open for "Data Mining with Weka"

We are now pleased to announce that the full Data Mining with Weka course is open for enrolment:

 

http://weka.waikato.ac.nz

 

The course will start on 9th September and extends over 5 weeks. It features:

 

• Video lectures
• Practical activities and online assessment
• Extracts from Data Mining (3rd Edition): Witten, Frank & Hall (2011)

 

  You can also follow the course via Twitter or the blog:

 

https://twitter.com/WekaMOOC

 

http://wekamooc.blogspot.co.nz/

 

 

Filming the WekaMOOC video trailer

Filming the video trailer next to the Gallagher Academy of Performing Arts:
 

Press Coverage

Some WekaMOOC coverage from the local newspaper:


http://www.stuff.co.nz/waikato-times/news/8942061/Online-courses-a-game-changer


"months out" is bit misleading; watch for announcements in the near future.

Recording environment

Some more photos from the WekaMOOC recording environment.

Acoustic foam to improve audio quality

Lighting

Recording preparation

We've closed the beta test and are studying the participants' feedback.

Here are a couple of photos of the recording environment for WekaMOOC:

HD Camera.

 

Microphone.

Beta test is closing on 24th April ...

We will be closing the beta test of the Data Mining with Weka MOOC on Wednesday April 24th. We will be studying the feedback to inform the delivery of the full MOOC later in the year.

Beta test of Data Mining with Weka will close next week ...

Thanks to everyone who is participating in the beta test of WekaMOOC. We plan to close the beta next week so we can concentrate on preparing the full course for later in 2013.

Filling in the survey at the end of the lessons will help us improve the full version of Data Mining with Weka.

Beta Test of "Data Mining with Weka"

The beta test of Data Mining with Weka is now ready to go. It is a closed beta test so only those with specific invitations can participate. Soon after the beta test is concluded we will open registrations for the full Data Mining with Weka MOOC. You can follow updates on the WekaMOOC twitter feed.

We hope to learn about the best way to present the material and configure the Google CourseBuilder platform so that the full MOOC, later this year, can run smoothly.




Screenshot from a video lesson in Data Mining with Weka