IANAB, but lately i have been thinking a lot about evolution and how it changes a species. The context i am considering is social evolution, but i think the rules of biological evolution apply. From what i understand, Reindeer have learned nothing from the event in general. In the localized area, a bunch of Reindeer are out of the gene pool and any of their kin that are close enough to encounter their carcasses but not the the cause are going to freely associate the deaths with whatever the perception sensory suite allows. My assertion: Evolutionary "learning" occurs over a long period of time through strings of events with similar inputs and outputs.
Not an argument or rebuttal; your comment made me think about it. :]
Nah, I was just kidding.
Unless there are similar events occurring during a lot of time, making those reindeers that prefer staying more separated and in lower areas have a significant advantage, and these preferences are hereditary...
Lots of people of course resort to cereals or some dessert (croissant, waffles...) but eggs and bacon is really weird. A toast with something on it (olive oil, butter, jam, the aforementioned "pan tumaca") is way more common.
With further training I don't think it's possible: since those movements are not useful nor harmful, they will appear in winning and losing matches.
A possible solution might be including some distance traveled metric in the reward function...
Oh, in Spain we also have those "it's cold in winter, it's hot in summer" 5-minute long news every week, with the apparently mandatory interview with some locals...
We get that in the UK, complete with the use of Fahrenheit in the summer and Celsius in the winter just so that they can use as big or a small a number as they want to.
I wonder how intelligent it is, since I do not feed that much data to Google Calendar, just the things I know I'm going to forget. Almost all of those empty hours in my Calendar are actually busy.
Garbage in, garbage out. There's other people out there that put in too much data (can't remember the link, but a guy put in literally _everything_, including nutrient consumption)
As an amateur, I've always wondered if reinforcement learning could work with games where there are some probabilities in place (e.g. poker).
What happens when the action taken is a good one but the outcome is negative due to bad luck?
Absolutely. Q-learning has this capabilities and a shallow neural network was used back in 1992 to play backgammon, which has a lot of stochasticity.
See https://en.wikipedia.org/wiki/TD-Gammon
ETFs are just investment funds you can buy and sell in the stock market, with their own symbol and everything just like any other stocks.
Index (passive) funds track an index, instead of trying to outperform the market.
You can have passive funds that are not ETFs (do not appear as a symbol in the stock market, have to go through a bank or whatever to buy shares), and active funds that are ETFs (actively managed).
Presumably any ETF is tracking something, that you could define to be an "index" in a degenerate sense (even if it's the "companies Buffet thinks are undervalued index", or the "solid gold bars index"). Or is there a technical definition of index?
