9,967 reputation
23984
bio website moderntoolmaking.blogspot.com
location San Francisco, CA
age 28
visits member for 3 years, 10 months
seen 6 hours ago

Nov
23
awarded  Good Question
Nov
20
awarded  Guru
Nov
17
comment R - glmnet - cross validated - AUC
@abhi: If thetype.measure function you're using is auc, then cvm IS the auc. cv.glmnet fits a whole sequence of models, and will report the auc for all of them. The max of the cvm sequence is the best model's auc. Please take some time to read the documentation for glmnet, it's very good.
Nov
16
revised R - glmnet - cross validated - AUC
added 1 character in body
Nov
16
answered R - glmnet - cross validated - AUC
Nov
11
awarded  Necromancer
Nov
6
comment Please explain the waiting paradox
@Glen_b That's a really good idea,hah!
Nov
5
comment Please explain the waiting paradox
I think with busses specifically there's an additional process where a late bus becomes later as passengers cram onto it, and the empty bus behind it eventually catches up (but remains empty). =D
Nov
5
comment R: partial dependency plots from GBM package. Values and y-axis
@woodland_creature: The plot is basically showing you the model's view of the world, so yes, a negative region of the plot indicates the model thinks values in that region will decrease the response. If my answer is sufficient, please feel free to up-vote it and/or mark it as accepted.
Nov
5
comment R: partial dependency plots from GBM package. Values and y-axis
@woodland_creature Yes, the new values are the modeled values of the response variable, after accounting for the average effects of the other variables. I'm not super familiar with the gbm.plot function but it doesn't look like there's an option to return the plot data. It looks like the function's written entirely in R, so you could copy the source code and modify it to return the plot data. Type: gbm.plot into your console.
Nov
5
answered R: partial dependency plots from GBM package. Values and y-axis
Oct
30
comment How to analyze elastic net fitted model coefficients
@jmnavarro Sounds good!
Oct
30
revised How to analyze elastic net fitted model coefficients
added 1117 characters in body
Oct
30
comment How to analyze elastic net fitted model coefficients
@jmnavarro I see your updates to the question, and appreciate them. See if you can distill your script down to the essence of the problem. Can you reproduce it with a subset of the columns from your original dataset? Can you reproduce it with a subset of rows from your original dataset? Can you reproduce it with a simulated dataset in a fresh R session? It's hard for us to help you debug your code without being able to run it on our own machines =D
Oct
30
revised How to analyze elastic net fitted model coefficients
added 1117 characters in body
Oct
30
comment How to analyze elastic net fitted model coefficients
@jmnavarro Please post an example matrix of binary data that generates more coefficients than expected. Until you do so, I (and everyone else on this site) will assume you just have a bug in your code.
Oct
30
answered How to analyze elastic net fitted model coefficients
Oct
30
comment How to analyze elastic net fitted model coefficients
@jmnavarro You need to post more than an excerpt. You need to post a data sample and code that reproduces the problem EXACTLY.
Oct
29
comment Data APIs/feeds available as packages in R
@robingirard If you do, add it to the Web Technologies and services CRAN task view!
Oct
28
awarded  Popular Question