- #1
tlonist
- 1
- 0
I have an expected value problem where z is a desired expected value and I want to reach and x is an amount I can vary.
There is a probabilty of success based on a logistic function ρ(x) with a reward of λx and failure with a probability of (1-ρ(x)) and loss of x. I am trying to solve for the correct value of x to reach an expected value z.
So:
[tex] z = p(x) \lambda x - (1-p(x)) x [/tex]
[tex] z = \frac{\lambda x}{1+ e^{-a-bx} } + \frac{x}{1+ e^{-a-bx} } -x[/tex]
I tried solving in Matlab but it says there is no explicit solution and I haven't been able to solve by hand.
What would be the next course of action to solve this? Is there a way to simplify?
There is a probabilty of success based on a logistic function ρ(x) with a reward of λx and failure with a probability of (1-ρ(x)) and loss of x. I am trying to solve for the correct value of x to reach an expected value z.
So:
[tex] z = p(x) \lambda x - (1-p(x)) x [/tex]
[tex] z = \frac{\lambda x}{1+ e^{-a-bx} } + \frac{x}{1+ e^{-a-bx} } -x[/tex]
I tried solving in Matlab but it says there is no explicit solution and I haven't been able to solve by hand.
What would be the next course of action to solve this? Is there a way to simplify?