Use the stated preference data to estimate values of time by market segment

[bhargavasana]

This is a follow-up from #355.

Use the stated preference data to estimate values of time for each of the four market segments. All else equal, using linear coefficients on time and cost is easier to interpret than log transformations. Log transformations make the implied value of time variable, which is not the worst thing, but harder to explain. Cost coefficients segmented by income should be explored, consistent with the existing model (i.e., the hypothesis being that higher income travelers are less sensitive to cost than lower income travelers).

[rickcurry]

Exploring rental car modeling strategy for visitors