How to pronounce“mu0”in English

So mu0 is x times beta 0 and so forth.

Let's estimate mu0 of x using a forest, again, forests are okay.

We could try estimating mu0 from data.

So here kind of mu0 is the expectation of y among controls conditional on x equals x.

So you could learn mu0 by learning to predict y from x among our controls or similarly learn mu1 .

tweaks to it but mu0 does effectively that it does those roll outs and those murdering in in a learned

How did you get these mu0 estimates?

if you look at alpha zero or like even mu0 if you look at that work you know there it's nice reinforcement learning

and then you started getting this thing which is kind of like deepmind's um mu0 on steroids self the ultimate self play

Here, the truth tau is the expectation of mu1 and mu0 .

And whatever shown so far is that if you were to do this using the true mu0 and true mu1 , then this will be

Then one could use these predictions as estimates for the function we care about, mu0 and mu1 .

You might have hoped for better, right, because I just started by writing tau of x is mu1 of x minus mu0 of x.

How good is this estimated average treatment effect by taking the average difference in predictions from them, you want to make some mu0 functions.