Weighted values from correlation coefficients

by Snupas   Last Updated July 19, 2017 17:19 PM

I've been pressed to find a similar existing example of what I'm trying to do, but maybe i'm not phrasing it incorrectly.

Essentially, I want to do a dynamic pricing model. Create a formula where i take the correlation coefficients of various variables related to pricing, and then when calculating a certain price, increment the price using a combination of the other variable value and its correlation coefficient to price.

Essentially creating some sort of weight value from the correlation coefficient that affects the calculations(the weights would get updated as the correlation coefficients get updated, as new datasets come in)

So if you have the variables : price, distance, timerange1(binary),timerange2(binary)

price = initial price + distance*correlation + timerange(0 or 1)*corr +timerange2(0 or 1)*corr

The operators in there are placeholders just to get the idea across, I was wondering if there is an established way of doing something like this, or the closest thing I could base this off of?

Thank you.

Related Questions

Averaging correlation matrices among subjects

Updated August 29, 2018 07:19 AM

Preliminary analysis for canonical correlation

Updated February 03, 2018 19:19 PM