20bits - Latest Comments in The Dangers of Genetic Optimization

Re: The Dangers of Genetic Optimization

jfarmer — Fri, 08 May 2009 18:49:51 -0000

My point was that you didn't learn anything about your customers that can be applied in a different tactical situation. You don't know anything about their motivations, their interests, or their state of mind.

So, although you get one result (increased revenue) you also lose something of value (verifiable knowledge about your customers). There's a tradeoff there, and perhaps you're willing to make it.

Re: The Dangers of Genetic Optimization

Someone — Thu, 07 May 2009 17:36:54 -0000

"There's no learning"? From a business point of view, there is. You found a way to increase revenues with little effort. That trick is worth remembering.

Re: The Dangers of Genetic Optimization

jfarmer — Sat, 06 Dec 2008 20:32:11 -0000

Greg,

Very cool. I'll keep an eye out.

I'd love to meet up some time. You're in SF, right? Are you ever down in the peninsula?

Re: The Dangers of Genetic Optimization

gregdingle — Sat, 06 Dec 2008 18:23:32 -0000

Hi Jesse, the black box aspect of genetic algorithms can definitely be a drawback. I've just launched genetify as an open-source project.

http://github.com/gregdingle/genetify/wikis/home

I'm hoping people will incorporate different testing and optimization algorithms into it -- ANOVA, bayesian, GA -- so you can pick one to suit your purpose.

Re: The Dangers of Genetic Optimization

Jamaal — Fri, 05 Dec 2008 13:01:14 -0000

Or can you apply it properly to this one? The world isn't linear, and combinations still matter.