November 2013 – Page 2 – DECISION STATS

Tails -an OS for Privacy

Just came across Tails.

https://tails.boum.org/

amnesia, noun:
forgetfulness; loss of long-term memory.

incognito, adjective & adverb:
(of a person) having one’s true identity concealed.

Tails is a live system that aims to preserve your privacy and anonymity. It helps you to use the Internet anonymously and circumvent censorship almost anywhere you go and on any computer but leaving no trace unless you ask it to explicitly.

It is a complete operating system designed to be used from a DVD, USB stick, or SD card independently of the computer’s original operating system. It is Free Software and based on Debian GNU/Linux.

Tails comes with several built-in applications pre-configured with security in mind: web browser, instant messaging client, email client, office suite, image and sound editor, etc.

Movie Review- Thor 2

This is Hollywood at it’s intellectual best. Natalie Portman, geeky queen of hearts ( since V for Vendetta and Star Wars 1-3) reunites with beefy actor Hemsworth (he of the Star Trek cameo ) , fellow Oscar winner Anthony ” The Hannibal” Odin Hopkins, and a full array of CGI to bring the old Norse God some real new meaning. Natalie is beginning to look a bit frail though.

Very entertaining and very light on brain movie. Unfortunately for Thor, Loki steals the show every time. Loki is a much better actor too and his presence tells us this will not be the last we see of feuding brothers

Oh yes, Asgard is lovely in this time of the year.

Clear Screen in #Rstats

Just use Ctrl + L to clear your #rstats screen.

Try it. It works like a clear screen

RapidMiner takes things to the next level

I have watched Rapid Miner for quite some years including the R -extension, interview with founders , one of the first marketplace for algorithms (or extensions to its statistical software) and use in sports analytics has been much in the news lately.

They got funded, revamped the website , changed the name from Rapid-I to Rapid Miner and are now announcing version 6 of their flagship software soon.

http://www.zdnet.com/rapid-i-gets-funded-re-brands-as-rapidminer-7000022757/

A well-kept secret of the Analytics/Data Mining world may get some of the spotlight now, with a cool $5M in its pocket.

a successful $5M Series A funding round, with participation from European firms Earlybird Venture Capital and Open Ocean Capital (the latter firm having a strong pedigree from the team behind the MySQL relational database).

It has easily been the first open source statistical tool with Visual Programming ( something R is still yet to have despite efforts by RedR, Analytic Flow et al) and more importantly has a huge stack of enterprise clients.

http://rapidminer.com/products/rapidminer-studio/

RapidMiner 6 will have brand-new templates for churn reduction, sentiment analysis, predictive maintenance and direct marketing. A data analysis toolbox has never been more user-friendly or more powerful.

But best of all- they get a much easier training academy in place, and I am personally going to finally master it ( even though I have played a bit with it before

I do hope they make a MOOC (since the software is open source and free to download – how about some very easy to do self learning online tutorials)!

http://rapidminer.com/learning/training/

Introduction to Data Mining and Predictive Analytics with RapidMiner Studio and Server, December 3^rd and 4^th.
This course is a two-day introduction to the foundations of data mining, business analytics, and RapidMiner software. Participants will gain a complete understanding of how RapidMiner Studio and RapidMiner Server work and are used.

This course is the perfect preparation for the Image Mining training course.

Foundations of image processing, analysis and mining with the “Multimedia Mining-Image” (MUMI-Image) extension, December 5^th and 6th.
This course is a two-day training on the foundations of image processing, analysis and mining with the “Multimedia Mining – Image” (MUMI-Image) extension for RapidMiner. After this training course, participants will have a complete understanding of how image mining analysis can be performed within RapidMiner Studio and Server, combining image processing techniques with the available data mining methods and data processing capabilities. Practical exercises ensure that the participants will be able to perform their own image analysis at the end of the class.

Blogger Disclosure- Rapid Miner has been a sponsor of Decisionstats.com for several years . I also like the software a lot!)

SAS Institute continues to lose training revenue as WPS clones use “language of SAS” loophole

Apparently, in Asia companies have been able to offer and advertise SAS Training , by inserting the language of SAS cleverly, hiding the disclaimers in the website map maze, an reassuring students that they can get jobs even in bonafide SAS Institute licenses client locations by defacto learning “the SAS language” without paying for the expensive SAS licensing.

This marks a double blow for the Institute , as one one hand WPS licensing erodes its margin by competitive discounting. On the other hand the lucrative SAS Language Training (and publishing) market is decimated in emerging market economies by SAS language clones.

I write this in dismay as I was one of the original authors of this article which was even referred to in the WPS judgement

http://en.wikipedia.org/wiki/SAS_language ( the edit history is even more fascinating if you want to see it!!)

Students, if you wanna learn SAS or language of SAS – do refer to this.

http://support.sas.com/training/

Anything else is a legal grey area compounded by the curious shyness of SAS legal to defend a 40 year old venerated analytics brand, by ceding ground through out the world and across domains

A simple analysis of Web logs in SAS Online Doc would prove how this is exploited even by sellers of SAS language clone software.

As the difference in SAS and WPS pricing is compounded on server licenses, companies in Asia have found the best thing to do is get a WPS server license and in fact offer it for training (or analysis) like a time shared cloud solution to multiple customers at the same time

It is an interesting thing to watch- because SAS Institute remains one of the lost holdouts on the West Coast to the Stanford mafia.

unrelated picture – a famous Bollywood movie

Interview Joseph Eapen CEO RMInsights #rstats #shiny

Here is an interview with Joseph Eapen, CEO of RMInsights an exciting data science company that have been using R for their work in providing decision support for the Entertainment Industry (do not confuse them with rminsight- which is almost the same url but without the s)

I found some of the work in applied data science cool enough to request an interview and they were gracious to respond at short notice.

Ajay Ohri (AO)- What are the innovative steps , products services and initiatives that you have been executing at rminsights.net

Joseph Eapen (JE)– Cinema: RMI has launched Cinema Audience Measurement with focus on Audience Appreciation. As you know, film making is an expensive and time-consuming exercise; to add to it, the audience sit through a movie, with great expectation, hoping it is worth their time, effort and money. Even though Box office collections is a yardstick of movie revenues, it is still far from any audience appreciation measures. CAM through http://screenratings.org provides a scientific tool that will aid you with movie appreciation scores and help you opine to aid others in choosing a movie to watch this week.

Television: Apart from CAM, RMI focuses on Data Analytics, especially forecasting techniques using R and Shiny and has developed a TV ratings forecaster as a Service.

Data Collection: Designs Mobile based Custom Surveys using ODK (OpenDataKit –Build, Collect, Aggregate)

AO- What are some of the applications and products that you have been developing using R language and R Shiny Applications

JE- TV Ratings Forecaster as a Service is the product we have developed using R and Shiny. It uses Professor Rob J Hyndman’s Forecast package that provides methods and tools for displaying and analyzing univariate time series forecasts including exponential smoothing; using the Holt-Winters HW approach.

The user has to simply upload a csv file containing historical TV ratings of any TV channel(s), usually 48 time-segments for each day, decide on the forecast horizon and the system predicts and plots the forecast.

AO- What is your vision for research and analytics going forward ?

JE- We are all at an interesting juncture, almost everything we do is captured at a transactional level and the so called Bigdata will be accessible for analysis. So in future, I see that we will need research to bridge gaps in this transactional data, so a large role in decision making will be played by analytics more than research (as we know it now). Both will co-exists, but data mining and analytics will take center stage.

AO- Describe your career journey from students day to CEO

JE- After I finished my engineering (Electronics & Telecommunication) in Mumbai, I went to Dubai in search for a job. Within 10 days, I saw a job vacancy in the local newspaper for a Project Engineer at a leading Market Research company – PARC (A Gallup affiliate). I applied and after 2 rounds of interview, landed the job. The company had just signed a JV with AGB Italia (an Television Audience Measurement company) to conduct peoplemeter based TAM in the Gulf region, starting with UAE. The JV was called AGB Gulf. During the expansion phase of the project I was send to Milan to test and initiate the purchase of peoplemeters for the expansion, and trained at their headquarters in Switzerland. Soon I was promoted as Head of AGB Gulf, and ran it for 10 years.. During this time I led the service through an successful audit by the Industry.

Following this I was instrumental in the launch of TGI (Single Source from BMRB) in the Gulf and LEVANT markets, again through a JV between PARC and BMRB in UK. The JV was called TGI Arabia.

This was the first 13 years of my career. Post that I joined as Head of Research & Development at MediaEdge:CIA (A WPP Company) in Dubai. After 3 years, I returned to Mumbai and joined aMap (An overnight Television Ratings agency, competing fiercely with TAM Media Research – A Nielsen/Kantar Company) as Director and promoted to CEO within 11 months.

After 3 years with aMap, I was appointed as CEO of MRUC (An industry body that issues IRS – world’s largest readership study, in India)

Post MRUC, I joined RMI as CEO… and the journey continues…

AO- What advice would you give to aspiring data scientists

JE- As much as, there will be large amount of data to mine and lot of techniques available to make sense of it all, it may get overwhelming and we should not lose track of the purpose of mining, that is, application of the learning and understanding to enhance our life and the world around us. It should not be – Exciting to Mine and Good to Know… It should be way beyond that.. totally purpose driven and put to use, aggressively

AO- What do you think are the criterion that companies should take into factor while outsourcing research and analytics (apart from cost)

JE- Apart from cost, the key factor should be turn-around speed and accountability. Not only should they be fast & meticulous, but how much they can stand behind their work and see that it serves the purpose, it was commissioned for, in the first place. So you see it goes beyond ‘just doing a good job’

About RMInsights

We have studied human nature and behavior for more than 20 years. Our reputation for delivering relevant, timely, and visionary solutions on what people around the world think and feel is the cornerstone of the organization. We employ many of the world’s leading scientists in management, economics, psychology, and sociology. We study market conditions in local, regional, or national areas to examine potential of a product or service. We help companies understand what products people want, who will buy them, and at what price.

Joseph Eapen
CEO

http://www.rminsights.net/