Coursera is not a MOOC

Consider the hypothesis above and consider the facts below

massive open online course= Mooc

but coursera is not Open

1416322284676

  1. You cannot share Coursera content on Youtube or any non Coursera- this effectively ties in people to the website
  2. The Website does not use any social login or OpenID
  3. There is enough evidence of cross selling including branding, selling books and now 50$ certificates
  4. There is more vigor to get easily monetization course curriculum than to get subjects that are relevant to society ( eg job training for laid off workers in rural USA or for farmers in Africa or for textile mill workers in Bangalidesh)
  5. No effort in localization and customization or translation despite multiple avenues for machine learning enabled translation APIs (cough cough Machine learning= ML)
  6. Prof ML insists on using Octave and not R. Everyone knows Sheldon Cooper is more cool than Howard Wolowitz, yet Prof  ML  will say – it is easy to prototype using the latter and not the former. Plus he does not really update his slides too much
  7. With a 5% pass rate you actually dont care about the (students) education as they are just (money) bricks in your (fire) wall
  8. Blog posts are shared but data on usage and engagement is not shared openly or through an API
  9. Lame efforts at open sourcing the platform
  10. Lamer efforts at keeping clutter off the course search page ( Hint- Dont use Ctrl + F)

All images sourced from http://boards.4chan.org/b/

1416314583026

 

Revolution Analytics likes FOSS and becomes seriously awesome

and I mean Free and Open Source (FOSS not Fuss)

Hopefully teachers in academia can start using the flavor of R that will help their students get more jobs

Hopefully analysts in industry can start using this flavor of R without worrying on costs or vendor lock in

It’s all here folks and its seriously awesome

http://projects.revolutionanalytics.com/

 

Screenshot from 2014-11-18 19:24:02

Screenshot from 2014-11-18 19:16:33

Screenshot from 2014-11-18 19:16:03

Wishlist for 2015- A Combined Corporation with RStudio and Revolution Analytics folks (   product synergies are awesome too 😉

RAnalytics

How to be a good blogger

1) Be Honest. Internet and Google are good at recognizing fakes

2) Always keep an eye on web analytics to receive feedback on what works and what does not work

3) Categories and Tags help people and search engines find your content easily

4) Trust in God but use Spell Check.

5) Interesting blog beats non interesting blog

6) Curiosity and Utility are the two things you need for a blog.

7) Artists steal. Bloggers get inspired. Search Engines catch thieves. So modify when you get inspired. Use original words

8) Benchmark with the best blogs in your subject. Read others. What does your blog aspire to do?

9) People who know CSS and HTML are better at changing blog layouts. Or people who are better at finding WordPress themes.

10) Plan your writing and always a few drafts handy.

11) Be regular in your passion for blogging. If you are not passionate about doing a good job out of it, you will NOT do a good job out of it.

12) Okay I have a superb blog. Where do I publicize without looking like a spammer?

Ajay Ohri in Wired – An App Store for Algorithms

I got featured in a Wired article recently- this is one of my very old ideas- basically an App store for Algorithms.

I  briefly advised two startups in this space (but no longer do)

Klint Finley took time to shoot me some questions and you can read the final article here.

http://www.wired.com/2014/08/algorithmia/

I have now been featured in Wired , ReadWrite Web and a member of Star Trek has reblogged me on Tumblr! Geek heaven and I owe it all to the readers of Decisionstats.com!

 

The Silence of the Data Science Lambs

While courts, politicians, activists , spies and even corporate leaders have spoken or ducked on the question of whole sale data collection by NSA, one group that is both in the thick of action as well as conspicuous by it’s silence is the data scientist community.

ill

While one prominent open source member of R community spoke out against analyzing the data leaked by Wikileaks ( an admirable stand given his background) no one seems to be perturbed to be speaking on analyzing data belonging to fellow citizens and the world at the same time. ( see-WHY I WILL NOT ANALYZE THE NEW WIKILEAKS DATA)

Now Data Scientists and the Intel Communities have long worked together and one of the SAS Institute’s solid cash cows remains its strangehold on intel analytics 😉 http://www.sas.com/resources/brochure/government-intelligence-community-overview-brochure.pdf) , what is perplexing is the deafening silence regarding the violation of Fourth Amendment rights of American citizens domestically and abroad  (see http://en.wikipedia.org/wiki/Fourth_Amendment_to_the_United_States_Constitution) and the active collusion by primarily data scientists in this

The right of the people to be secure in their persons, houses, papers, and effects,[a] against unreasonable searches and seizures, shall not be violated, and no Warrants shall issue, but upon probable cause, supported by Oath or affirmation, and particularly describing the place to be searched, and the persons or things to be seized

Is not your email , your social media, your mobile data a part of your person/house/paper and effect. Does Atlas need to shrug and do data scientists need to say enough is enough to stop this blatant misuse?

No- because the well funded NSA and DOD budgets will always be more than the conscience of a few data scientists. There will always be hackers for hire, and the people shall be led by a sheep.

The Numerati- or the  numerically enabled technological elite data scientists are as culpable as the agencies using them. This can be addressed by lawsuits against compliant statisticians and data miners as well as they are the ones enabling violation of fourth amendment rights. Countries like India have chosen to feed off this data trough and countries like China have chosen to create their own walled off internet instead. It is American data scientists alone who can help guide their Congress to Sanity. The timing is pertinent as Congress debates amending the Foreign Intelligence Surveillance Act

” the proposed changes would not touch the agency’s abilities overseas, which are authorized by Executive Order 12333, a Reagan-era presidential directive. The administration has declassified some rules for handling Americans’ messages gathered under the order, but the scope of that collection and other details about how the messages are used has remained unclear.”

 

( see – http://www.nytimes.com/2014/08/14/us/politics/reagan-era-order-on-surveillance-violates-rights-says-departing-aide.htmlhttps://www.documentcloud.org/documents/836235-ussid-sp0018.html and http://www.nytimes.com/2014/07/25/us/politics/senators-bill-is-stricter-on-nsa-than-houses.html)

Are you a data scientist who wants to help out? Help the ACLU educate Congress ( https://www.aclu.org/national-security/fix-fisa-end-warrantless-wiretapping )on the proper way to dispose off private data, and anonymize the data already collected .

 

Otherwise the next generations will be born in an age where every move recorded by CC cameras or wearable computing devices will be mined by corporations for ads and governments for threats.

The last word on this was said by a wise old White Man ( in an age where Wise Old White Men are no longer fashionable or even correct)

The tree of liberty must be refreshed from time to time with the blood of patriots and tyrants–Thomas Jefferson

Maybe he was referring to this tree

tree

 

 

The first Decisionstats.com Intern

The following certificate is awarded to Chandan Routray (https://www.linkedin.com/in/chandanroutray) , a 2yr student at an IIT who learnt all this

and wrote some of  this at https://python4analytics.wordpress.com/– the following certificate to show he was tested as a potential scientist and showed great promise in executing his task!

chandan