I recall reading about a paper in SciAm or American Scientist a couple of decades ago, where they had trained a ML model to predict regional conflicts or civil wars. The main input was scarcity of food, mainly through price IIRC.

They trained it on historical data up to the 90s or so, and had it predict the "future" up to the time of the article. And as I recall it did very well. They even included some actual near-future predictions as well which also turned out pretty accurately as I recall.

Which I suppose isn't a huge surprise after all. People don't like to starve.

▲

schainks an hour ago | parent [-]

Link?

	▲	zweifuss 37 minutes ago \| parent \| next [-]
		https://homerdixon.com/wp-content/uploads/2017/08/Environmen...
	▲	magicalhippo 37 minutes ago \| parent \| prev [-]
		My memory isn't good enough to recall the name of the paper, however doing some searching I see the field has not stood still. Here[1] is an example of a more recent paper where they've included more variables. A quote from the conclusions: The closest natural resource–society interaction to predict conflict risk according to our models was food production within its economic and demographic context, e.g., with GDP per capita, unemployment, infant mortality and youth bulge. [1]: https://www.mdpi.com/2071-1050/12/16/6574