Let’s Co-Create A New Vision for SMAC Convergence


We are now writing the introductory chapter in TechAmerica’s research examining the nexus of social, mobile, analytics, and cloud.  CSC is hosting a #SpeedIdeation event with an open discussion on the changing role of government, the meaning of national security, and our entire approach to intelligence collection and law enforcement.
Your comments will help shape the research and we will, of course, acknowledge your contribution in the final report.

#SpeedIdeation CrowdChat Sep 22 – 3pm CEST, 9am EDT

This coming week we’ll start one hour earlier. Soren and Thomas will lead a discussion on public WiFi for business purposes as discussed on Twitter.

The two main hosts will publish more information soon.

Chat will be as usual here: https://www.crowdchat.net/SpeedIdeation



Next #SpeedIdeation Event on Monday, September 15

Looking forward to global creativity again.

Thanks to @jerryaoverton for hosting http://www.via-cc.at/91mqh


Software Engineering for the Data Scientist

I came up as a software developer and only recently have I gotten into data science. Software engineering is in my bones, but for many of my colleagues, software engineering is a bit of a mystery.  That’s a problem because it affects productivity.  Data scientists need software engineering skill — just not all the skills a professional software engineer needs.  So what are the essential software engineering practices needed in data science?
In this #SpeedIdeation I’m calling on all software engineers, hackers, and data scientists to brainstorm on the best practices needed to write solid code in data science.

CameraTracker Interface

Das Interface ICameraTracker ist nun an den Renderer angebunden und eine statische Klasse implementiert, welche das bisherige, starre Geradeausschauen abbildet. Dabei ist mir aufgefallen, dass letzterer für die Kamera leider keine Winkelausrichtung unterstützt, sondern nur ein “looking target”.

Die Idee dabei ist, dass die Kamera sozusagen an Position (0,0,0) gelegt wird und ein virtuelles Objekt existiert, welches irgendwo im Raum positioniert wird. Die Kamera wird dann angewiesen, auf dieses Objekt zu zielen. Der Einfachheit halber ist es sinnvoll, hierfür einen begrenzten Raum von (-1,-1,-1) bis (1,1,1) aufzuspannen, in dem das Objekt positioniert wird. Der Vorteil hierbei ist, dass die Blickrichtung im Gegensatz zu Eulerschen Winkeln (wo es von der Ausführungsreihenfolge abhängt) eindeutig ist.

@Anton: Ich könnte jetzt die Winkelangaben umrechnen, nehme aber an, dass es effizienter ist, wenn das bereits auf Deiner Seite passiert. Ich habe daher das Interface in diesem Parameter angepasst. Die Details Deiner bisherigen Berechnungen habe ich leider nicht durchdrungen, vielleicht kannst Du es soweit optimieren, so dass Du direkt die Raumkoordinaten ermittelst statt der Winkel?

#SpeedIdeation progress

After some days of vacation and business travel we’re now preparing the next session. In the meantime we did get some traction in the project(s) started and also received additional ideas.

I will post some updates over the weekend – next session date will also be published until Sunday


DBAR / TrainInspect MoM 2014-09-05

As discussed, guys:

– Please tag all your posts with “DBAR” to ensure they’re listed on the project page.


Join #SpeedIdeation 2nd event today

In less than one hour @JerryAOverton will answer questions to the paper published last week.

Send all your questions during CrowdChat – and keep going with additional ideas


2nd #SpeedIdeation: Simulating New Business Innovations

We’ve created a new method of business model simulation and used it to simulate the market disruption of Blockbuster by Netflix.


Click here for the interactive simulation: https://plot.ly/~jerryaoverton/4

The simulation anticipates that the Netflix model can become profitable with much less investment than is required for the Blockbuster model; and it anticipates that the Netflix model has a greater overall revenue potential than the Blockbuster model [Des]:


Table 1: The observed performance of Blockbuster and Netflix. Source [Des]

The simulation predicts a period of similar performance, followed by a period where Netflix far outperforms Blockbuster, followed by another period of similar performance.  Gross profit for Netflix and Blockbuster shows agreement between simulation and actual performance:


Figure 3: Gross profit of Blockbuster and Netflix from 1998 to 2008 [Top]

The method was useful in simulating real innovation, it made predictions matched observations, it produced real insight, and was easy to build using R.

Want to know more?  You can reach me on Twitter: @jerryaoverton

[Des] Amit Deshpande, Analytics: Netflix vs. Blockbuster, http://bit.ly/1shL3XI

[Top] Top Accounting Degrees, Netflix vs. Blockbuster, http://bit.ly/WUCeZs

1st #SpeedIdeation CrowdChat ended successfully

Thanks again to everybody who helped prepare, run and of course to all of those who joined in and volunteered to “get stuff done”.

We’ll open the projects we decided upon within the next 24 hours.

CrowdChat stats:

Social Impact

Total Timeline Impressions 61,380
Total Views 751
Tweets 115
LinkedIn Posts 8
Facebook Posts 0
Peak Online Users 29


Total posts 129
Unique people from Twitter 23
Unique people from LinkedIn 5
Unique people from Facebook 0