p*******i
发帖数: 309 | 1 1. Variance reduction.
Antithetic variables, simulate U and 1-U, x=F(u), y=F(1-u), there is var[(F(
u)+F(1-u))/2] is less than var(F(u))
Control variate method, E[Y] is known, then Var(X+c(Y-E(Y)) is less than Var
(X)
2. How to generate pi,
3. any problem in a correlation matrix, A->B cor: 0.9 and B->C cor: 0.9
then A->B should not too small.
4. what is the so- called arbitrage in finance, and arbitrage with risk.
5. why people use BS do not use monte carlo simulation. Because BS is good.
6. wh | b***k
发帖数: 2673 | 2
F(
Var
pi=2*acos(0.0) or pi=4*atan(1.0)
【在 p*******i 的大作中提到】
: 1. Variance reduction.
: Antithetic variables, simulate U and 1-U, x=F(u), y=F(1-u), there is var[(F(
: u)+F(1-u))/2] is less than var(F(u))
: Control variate method, E[Y] is known, then Var(X+c(Y-E(Y)) is less than Var
: (X)
: 2. How to generate pi,
: 3. any problem in a correlation matrix, A->B cor: 0.9 and B->C cor: 0.9
: then A->B should not too small.
: 4. what is the so- called arbitrage in finance, and arbitrage with risk.
: 5. why people use BS do not use monte carlo simulation. Because BS is good.
|
|
