Re:Can Mathematics Unlock the Secrets of Terrorism?
by giardini on Wednesday March 25, 2009 @07:54AM
(#531711)
I'm surprised to see decision trees referenced: they are unsuitable for modeling real-world situations unless _every_ possible alternative can be determined aforehand, something simply impossible for terrorism (or for most human activity, for that matter).
[ Parent
| on HN ]
Re:Can Mathematics Unlock the Secrets of Terrorism?
by trapper on Wednesday March 25, 2009 @09:23AM
(#531772)
Possible scenarios can easily be abstracted. E.g. Building attack, car bomb, etc etc. They have the data.
Re:Can Mathematics Unlock the Secrets of Terrorism?
by giardini on Saturday April 25, 2009 @11:47PM
(#579641)
Certainly scenarios can be and are described but not all possible scenarios can be listed because they are infinite in number, And this is true in a non-trivial sense.
When you say "They have the data." I assume you are speaking of statistics. But
- These are rare events and statistical analysis is at it's limits,
- The past is no predictor of the future. Example: the introduction and continued enhancement of IEDs in Iraq.
Although he doesn't elaborate on my first point above, I recommend that you also read Nassim Nicholas Taleb's book, "The Black Swan".