–1 vote
asked ago in General Economics Questions by (300 points)
For a project I'm working on, I'm looking for a model that when put to the data has both local minima, of course, a global minimum.  The model is not what interests me, I'm going to apply a machine learning idea I've had, so the simpler the model the better.   But having non-trivial and difficult to find global minimum is the most important point (although obviously, you don't have to actually know what that is).

1 Answer

+1 vote
answered ago by (460 points)
selected ago by
Best answer
Not sure if this is what you want. Check out https://www.sciencedirect.com/science/article/pii/S0304407613000419.
The model is very simple: data(tau)=(1-a*(1-cos(omega*tau)))/(1+b*tau^c) where a,b,c,omega are parameters.
But the likelihood surface is very bumpy thanks to the cos function.
commented ago by (300 points)
perfect!  I will use.  Thanks.