Wolverine said:As an A2 level Biology student and one who's going to do biomedicine next year, i'd say protein synthesis via DNA's transcription and translation is the most ingenious thing i've learnt thus far.
Natural selection, speciation and all that is alright but usually a bitch to answer in the papers; this year we had an allopatric speciation question on male seahorses. feck sake.
Wolverine said:Aye, the males get pregnant; embarrased to say that i didn't know that.
Wibble said:Dead easy. Talk about allopatric speciation in seahorses in the way you would talk about any geographically isolated form of speciation. To get an A say that sympatric speciation may have also occured due to size specific breeding habits in some seahorse species attributed to males brooding the young. If you mention disruptive selection (where large and small individuals survive better than intermediate sized individuals) then your lecturer will give you an A+ and want to dry bum you.
spinoza said:When I went to Malta, the dive guide said "Ve have too zeahorzes *** are very beautiful in der cave".
I thought he meant two types of seahorses.
He did. And he also meant 2 seahorses.
In a cave the size of a large sitting room. Pitch dark. With coral, rocks and other crap.
Cue 6 divers bobbing around with torches trying to spot 2 creatures the size of my little finger.
We found them in the end though, but not before someone ran out of air...
spinoza said:I thought sympatric speciation was unproven.
nickm said:Paper. Not papers. Books and websites - not peer-reviewed, therefore it's pseudo-science.
Mihajlovic said:Peer-Reviewed & Peer-Edited Scientific Publications Supporting the Theory of Intelligent Design:
http://www.discovery.org/scripts/viewDB/index.php?command=view&id=2640&program=CSC%20-%20Scientific%20Research%20and%20Scholarship%20-%20Science
So, how many of these have you read?Mihajlovic said:Peer-Reviewed & Peer-Edited Scientific Publications Supporting the Theory of Intelligent Design:
http://www.discovery.org/scripts/vi...Scientific Research and Scholarship - Science
I thought sympatric speciation was unproven.
Peer-Reviewed & Peer-Edited Scientific Publications Supporting the Theory of Intelligent Design:
http://www.discovery.org/scripts/viewDB/index.php?command=view&id=2640&program=CSC - Scientific Research and Scholarship - Science
I thought sympatric speciation was unproven.
I fecking hate Science and Nature. Highest rated journals, but the quality of papers is so much worse than corresponding top journals (or in CS conferences) in particular fields. The machine learning/statistics part of the publication was nuts, no way in hell would have passed in top ML conferences. I guess the prize to pay for inter-science articles though (combine different fields, but be bad in some/all of them).The jury is in and it is now proven
https://phys.org/news/2015-12-darwin-puddle-species-emerge-geographic.html
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4700518/
I fecking hate Science and Nature. Highest rated journals, but the quality of papers is so much worse than corresponding top journals (or in CS conferences) in particular fields. The machine learning/statistics part of the publication was nuts, no way in hell would have passed in top ML conferences. I guess the prize to pay for inter-science articles though (combine different fields, but be bad in some/all of them).
I don't feel it provided too much value, but at the same time, much better articles have shown this happening in more primitive forms, and there is already a mountain of proof.
I think it is more that they need to be lite in specific sciences, in order to appeal to everyone. Additionally, they cannot go too much into details cause the reviewers might be totally clueless in that scientific discipline. While I get the point of them (and the popularity of them talk for itself), they look quite unprofessional to me (even those who are pure machine learning / computer vision are quite meh and have long given up on them).I bow to your knowledge of machine learning as I know next to nothing about it.
One issue with most journals is that their pool of reviewers come from a particular discipline. Biologists will have biology and multivariate biostats skills but probably no machine learning expertise for example.
When we first discussed this many years ago the proof was largely modelling agreeing with outcomes that made it possible but not really proven (if I remember correctly - which I may not).