I'd just like to say it should be complex statistics. But that's just me.
Even AlphaZero, which, iirc, is trained entirely using self-play, with no starting data from other players?
Of course the situation becomes rather interesting when you start training it against itself, but you're still fundamentally trying to find a good statistic to estimate your chances of winning.
Perhaps I am not using the right definition of “statistics”?
More generally I'd consider statistics to be the applied version of probability theory. Of course in this case the very thing they were trying to compute also fit the definition of a 'statistic'.
If you consider this is to be too broad, then keep in mind that it's simply better when you can apply the concepts and techniques from probability theory to more things.
It seems to me that there is a category of people who are eager to dismiss deep learning altogether and say "iT's JuSt stAtIStics" even though there is a good amount of evidence to show that it isn't the case. That isn't real science, it's human bias.
Wrong.
If you actually study signal processing, you will find out that CNNs aren't something magically rooted in something other than "statistical correlation machines." CNNs in fact work because they're used to calculate cross correlation!
"Convolving" in terms of a CNN is a misnomer, it's the same as calculating cross correlation in terms of signal processing.
> Wrong.
The best way to respond to this, somewhat humorously would be "Wrong".