Today: you have a biased coin which lands heads with some probability p∊(0,1), and you can toss it as many times as you want. You don't know p, though...
1/10
(Also, you'd like to do this efficiently, thank you very much.)
Let's start with a simple function: f(p)=1/2.
Q1: Can you?
2/10
Ha.
A finite automaton.
3/10
Q2: What about f(p)=1/7?
4/10
Remember: you don't know p, but this should work for all p!
Q3: What about f(p) = p²(1-p+p²)⁻¹
5/10
Q4: What about f(p) = √p?
6/10
7/10
Q5: for f(p) = 2p?
8/10
Q6: Would any of the results change if you were allowed *any* type of computation?
9/10
[Note: I haven't even mentioned how many coin flips are required, i.e., how "coin-efficient" those computationally efficient processes are...]
10/10