Regression analysis and human frailty, relationship between!
For lassiterfics:

You can train a neural net via regression analysis, but is that how a human mind works? I'm not sure. They say that nerves that fire together wire together and that might (or, for all I know, might not) be equivalent, but there's also short and long term potentiation and that seems to be important in the human brain and it doesn't, at first glance, look like regression. I mean, when we form memories, are we fitting a scatterplot in our head?

Of course, we fit scatterplots in our heads all the time. When we're trying to recognize patterns (which we do even when there are no patterns to be found), when we're trying to understand causation (which we're /so/ bad at), when we're trying to predict time series (we only consider the recent past and then mostly ignore new data), and when we're trying to induce/infer other relationships (I think David Hume had a rant or two about that). So our regressive scatterplot engine is a really, really flawed. Stands to reason it didn't need to be all that good when we were running across ancient African grasslands but now we need something better. We have to stop fooling ourselves, and ourselves are the easiest people to fool.

So regression analysis is a way of taking a powerful but buggy unconscious method and making it conscious. We can't juggle near as many variables this way, but at least we can correct for our well-known biases. Regression analysis helps us focus on the bits that matter, helps us resist the temptation to make those mistakes which we're so very good at tripping over, helps us gain perspective on the whole dataset. It's a way to, for a little while, act as if we're smarter and wiser than we really are. A way to, maybe, become less frail.