3 ms·
Just try a problem like this: https://stats.stackexchange.com/questions/302891/hyper-parameters-tuning-random-search-vs-bayesian-optimization https://stats.stac
by nonbel 8y ago
Just try a problem like this:
https://stats.stackexchange.com/questions/302891/hyper-parameters-tuning-random-search-vs-bayesian-optimization https://stats.stackexchange.com/questions/302891/hyper-param...
The space of possible configurations can be simply huge, but still you can usually find a reasonable (not optimal) one after only a dozen or hundred attempts.
>"single ionization events can kill a cell’s ability to reproduce (or worse)"
Maybe I don't know what you mean by single ionization event, if it is mutating a single basepair then I doubt it.
- jcims 8y agoRe: ionization i meant a single 'bit flip'. I've since found this article, the effect seems to depend heavily on where it happens: https://en.wikipedia.org/wiki/Point_mutation https://en.wikipedia.org/wiki/Point_mutation I think your other point is probably related to sparsity of viable vs. non-viable configurations and the steps needed to get between them. Of course that's complicated by the contextual interdependecies (are you in a mammal's womb or a submerged egg sac) but to your point you don't need to climb mount everest in a single step. Just get from one solution to another and the viable ones will tend to outweigh the less viable. Appreciate the help reframing it a bit. I'm still thinking it's too complex but at least there's light at the end of the tunnel. :P