Discover MakerZone

MATLAB and Simulink resources for Arduino, LEGO, and Raspberry Pi

Learn more

Discover what MATLAB® can do for your career.

Opportunities for recent engineering grads.

Apply Today

Thread Subject:
endogenous MLE standard error computation

Subject: endogenous MLE standard error computation

From: C F

Date: 15 Aug, 2009 07:11:01

Message: 1 of 2

Hi-

I'm using MLE to estimate a 5 parameter model, call them: (a,b,c,d,e.) Parameters (d,e) are a function of (a,b,c) so I've resorted to a grid search method where I maximize the likelihood function over (d,e) using fminsearch conditional on a specific value of (a,b,c). I use a grid method to pick specific values for (a,b,c) over a finite range. This seems computationally inefficient, and though I'm able to get my parameter estimates, I cannot get the bootstrapped standard errors I need because simulations simply take too long (days or even weeks) Two questions:

1) Is there a better way to run this type of MLE?
2) If not, is there an alternative way I can get standard errors for my parameter estimates?



Thanks,
CF

Subject: endogenous MLE standard error computation

From: Peter Perkins

Date: 17 Aug, 2009 14:17:00

Message: 2 of 2

C F wrote:
> Hi-
>
> I'm using MLE to estimate a 5 parameter model, call them: (a,b,c,d,e.)

These two statements
> Parameters (d,e) are a function of (a,b,c)

> so I've resorted to a grid search method where I maximize the likelihood function over (d,e) using fminsearch conditional on a specific value of (a,b,c).

seem incompatible. If (d,e) are a _function_ of (a,b,c), then it would seem that, given a particular (a,b,c), you need to compute the value og (d,e) as a function of (a,b,c), and then plug the resulting (a,b,c,d,e) into your likelihood function. Maximize that over (a,b,c), using standard methods. You described maximizing a likelihood over (d,e) and that doesn't seem right.

Perhaps I've misunderstood.

Tags for this Thread

What are tags?

A tag is like a keyword or category label associated with each thread. Tags make it easier for you to find threads of interest.

Anyone can tag a thread. Tags are public and visible to everyone.

Contact us