« first day (2110 days earlier)      last day (2882 days later) » 

Timothy Nodine
Timothy Nodine
1:09 AM
Currently working on answering stats.stackexchange.com/questions/214696 and would like a sanity check.
It looks like his error function gives either zero error or infinite error for any discrete distribution.
 
 
5 hours later…
Dawny33
Dawny33
6:18 AM
Can someone please help me with this regression question:
0
Q: What does it mean when the coefficient of the squared term is negative in regression?

Dawny33I am reading a research paper which models a regression model where the returns are regressed on the number of ad exposures. the equation looks something like this: $Returns = beta_1*nExp + beta_2*nExp^2$ nExp: The number of times the user looks at the ad. Returns: The revenue from the user ...

regression
 
 
3 hours later…
Stephan Kolassa
Stephan Kolassa
9:04 AM
@Dawny33 Robert de Graaf's answer datascience.stackexchange.com/a/11946/2853 is spot on. I commented a bit.
 
Dawny33
Dawny33
@StephanKolassa Ahh okay. So, the negative coeff. for the squared term means that the rate of increase decreases. Got it! :)
 
Stephan Kolassa
Stephan Kolassa
@Dawny33: exactly. Because the rate of change is the derivative... so the rate of change in the rate of change is the second derivative. And the second derivative of a second order polynomial is just $2\beta_2
$2\beta_2$. So the rate of change in the increase (the increase is the original rate of change) is negative.
 
 
14 hours later…
Halil Pazarlama
Halil Pazarlama
11:21 PM
hi all, i have a possible opinion-based question so i decided to ask here. do you say "random k-fold cross validation" to state that you selected the folds randomly? or is it implied, so is it enough just to say "k-fold c.v."? or is there a fancier term? :) thanks!
 

« first day (2110 days earlier)      last day (2882 days later) »