In evolved or engineered systems, the optimal amount of any manipulable quantity is rarely unbounded in the positive or negative direction.
-
-
Replying to @MakerOfDecision
2/n This is find as long as you're not optimizing blindly; in economics, non-satiation is frequently true over any *reasonable* range.
1 reply 0 retweets 1 like -
Replying to @MakerOfDecision
3/n But automated optimization, as opposed to humans who use gut checks, frequently finds value(s) to boundlessly maximize/minimize. Why?
1 reply 0 retweets 2 likes -
Replying to @MakerOfDecision
4/n Simplified models, partly. If something has increasing benefits in observed ranges, optimization leads to out-of-sample extrapolation.
2 replies 0 retweets 2 likes -
Replying to @MakerOfDecision
"Distance from known good value" can be a variable to optimize.
1 reply 0 retweets 0 likes
Replying to @CauseOfProblem
Sure, there are cases where it doesn't happen, but transforming variables like that isn't a proof of that point.
6:02 PM - 20 Jul 2017
0 replies
0 retweets
0 likes
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.