I am trying to fit a regression via maximum likelihood estimation, one of the regression terms involves $\beta_0e^{(\beta t)}$ where $t$ is measured in hours and has a range of 0 to 90 days.
The inverse of my hessian contains negative terms on the diagonal, which leads to a not interpretable standard error. Would changing the units of $t$ help in fixing this?
My initial thought is that if the term contributes a small effect, $\beta_0$ will have to be very very small to compensate for $e^t$ growing exponentially.
