In the online dating market of 2019 looks is all that matters. Because of these reasons:
1. Online dating mostly depends on looks, it's based on your pictures. Almost no woman reads your bio. Simply because people don't have time (this is another reason why new apps like tinder and bumble even don't give you enough spaces to write about yourself, because it doesn't matter at the end).
2. If a woman does not find an attractive man "interesting" after first couple of dates or so, she can easily find another attractive man with "interesting" personality. They do not have to go for an ugly male with interesting personality. Because women (even average/non-attractive ones) have an extremely large dating pool, men don't. For example:
Same sex (specially between men and men) is a completely different ballgame. The mate choice strategy of women is completely different from that of men.
In a free/unregulated sexual market, women are the selectors, not men. Women are biologically hardwired to select good looking males (irrespective of his money, personality or social status). Therefore ugly people don't get matches. Very simple.
No matter how many ways you try to engineer/fix the system, it's not going to work. Because the existing system is already doing what it is supposed to do.
Those who are wondering why Indian languages are a subgrouop of Indo-European languages, this chart might be helpful. It shows some of the many words that Sanskrit and Slavic both share.
Interestingly, the style of art and writing is very similar to that of Hokusai Manga. Which is considered as the one of the predecessors of today's Manga and Anime.
Why all these sightings are never accompanied with sonic booms? If something is moving that fast through the atmosphere, there must be a sonic boom. Or is there any way to suppress this? At least theoretically?
Yes, that's true. But in optimization domain the concept of "convexity" is understood in terms of set, not always from the 2nd derivative of a function. Because you might have search spaces where you are not able to differentiate the objective function at all. In those cases the "convex" means a "convex set".
1. It's theoretically impossible to guarantee a convergence to global optima using gradient descent if the function is non-convex.
2. The only way to guarantee is to start the gradient descent from different points in the search space or try with different step sizes if the algorithm only starts from the same point in the search space.
3. Also does "achieving zero training loss" mean the network has converged to the global optima? I used to know you will get zero training loss even if you are at a local minima as well.
I work on a project related to combustion and fuel efficiency and we need to run thousands to fluid mechanics simulations/FEA on the cluster computing facility in our university. Sometimes they go like weeks to finish, a powerful clsuter machine will definitely speed up everything.
Is there any technical paper on this somewhere on arxiv? I was trying to implement one of my own without all the enterprise abstractions and now I am stuck.
To folks who are complaining the trade-off between it's price and what it does offer: this dog is more about a robotic toy with animatronics similar to that of spot-mini than a cloud powered AI (i.e. siri/cortana/google). You are buying an AI powered hollywood level animatronic toy for 2 grands, I wouldn't say this is too expensive.
I think even if they had no rock formation or the whole layer was uniform, they would still deflect. It's similar to walking in a straight line blindfolded.
https://www.youtube.com/watch?v=JLVmBJEzXhk