Tag Archives: MIT

Statistics of Violence as Special Case of Fat Tails

A lecture summarizing the paper on violence with Cirillo to MIT and NECSI data scientists, mostly explaining to data scientists how to work with fat tails and only indirectly addressing Pinker by responding to those fooled by the illusion of drop in violence:

 

https://www.youtube.com/watch?v=R1Dm2ZYeA6U

 

Uploaded on Jul 10, 2015

A lunch discussion at MIT with computer science-data science researchers from MIT and NECSI. The talk is mostly about the methodology for dealing with fat tails, with application to violence.
We show which claims can be made, and which ones cannot be made, from data, and “violence has dropped” is not one of them. We do not focus directly on Steven Pinker’s popular science book (owing to some deficiencies), only indirectly as some people in political science made reference to it.
The paper is here: http://arxiv.org/abs/1505.04722

via: Facebook