A standard technique in many sciences, particularly social sciences, is something called “linear regression” and its more complicated descendants (multiple regression, structural equation modeling, hierarchical linear modeling, …). This article critiques the overly uncritical usage of such techniques:

Goertzel: Why Regressions Fail

In computer science and mathematics that is closely related, a major unsolved problem is whether a certain class of problems that can be solved by “fast” algorithms is equivalent to another class of problems that can be easily checked *given* a solution but where the most efficient algorithm for solving them is not known to be “fast.” This conjecture is called P = NP. In fact, most people think that the conjecture is *false*–that just because you can check a problem’s solution quickly, you can’t necessarily find a quick algorithm for doing the problem. This article explains the issue very nicely:

MIT: P vs NP

And since we’re talking about math, and in *Frontiers* we’ve recently taken a look at probability, I thought it’d be nice to point you all toward one of my favorite radio shows looking at the question of what randomness is:

Radiolab: Stochasticity

Enjoy.

### Like this:

Like Loading...

*Related*