Rank: 278 based on 254 downloads (last 30 days) and 8 files submitted
photo

Patrick Mineault

E-mail
Company/University
McGill University
Lat/Long
45.50872039794922, -73.58116149902344

Personal Profile:

http://xcorr.net

Professional Interests:
neuroscience

 

Watch this Author's files

 

Files Posted by Patrick View all
  • 1 - 5 of
  • 8
Updated   File Tags Downloads
(last 30 days) 		Comments Rating
07 Dec 2011 Screenshot Fit GLM with quadratic penalty Fits GLM with a quadratic penalty, determines hyperparams through cross-validation or evidence Author: Patrick Mineault generalized linear mo..., regularization, penalized regression, map, glm, linear regression 15 5
  • 5.0
5.0 | 2 ratings
02 Nov 2011 Screenshot Auto Gaussian & Gabor fits Fit 1D/2D Gaussian or Gabor to a curve/surface without start guesses for params Author: Patrick Mineault gaussian, curve fit, global optimization, gibbs sampling, surface fit, 2d gaussian 97 17
  • 4.8
4.8 | 5 ratings
22 Sep 2011 Screenshot Matching pursuit for 1D signals Performs Matching Pursuit on 1d (temporal) signals with custom basis Author: Patrick Mineault matching pursuit, basis pursuit, convolution, spectrogram, timefrequency analysi..., temporal signal 43 0
02 Sep 2011 Screenshot Fast B-spline class Fit, evaluate, differentiate non-uniform B-splines of any order - fast Author: Patrick Mineault bspline, pspline, spline, cubic spline, smoothing spline, interpolation 43 0
02 Jul 2011 Screenshot Boosted Generalized Additive Models (bgam) package Boosting for the Generalized Additive and Linear Models (GAM and GLM). Author: Patrick Mineault glm, gam, generalized linear mo..., generalized additive ..., boosting, maximum likelihood 33 4
Comments and Ratings by Patrick View all
  • 1 - 5 of
  • 23
Updated File Comments Rating
22 Apr 2012 Fit GLM with quadratic penalty Fits GLM with a quadratic penalty, determines hyperparams through cross-validation or evidence Author: Patrick Mineault

Either Generalized Linear Models by McCullough and Nelder or Generalized Additive Models: An introduction with R by Simon Wood.
26 Mar 2012 Auto Gaussian & Gabor fits Fit 1D/2D Gaussian or Gabor to a curve/surface without start guesses for params Author: Patrick Mineault

Ben, bootstrapping might work incorrectly if the Gaussian bump takes only a few pixels. In general, MCMC should be more reliable. The only reason I would use bootstrapping over MCMC is that MCMC is mathematically complicated and it's a bit tough to explain in the methods section of a paper.

Daniel, it's difficult to say without having access to the data you have. It's possible that there's a bug in the code that only appears for very specific datasets. Try to boil it down to the simplest example that shows the bug and send me the code to replicate the problem.
20 Feb 2012 Auto Gaussian & Gabor fits Fit 1D/2D Gaussian or Gabor to a curve/surface without start guesses for params Author: Patrick Mineault

Ben, look at the r2 values that Matlab gives you. If it's 1, then it's a perfect fit, if it's 0, then not. The other cue is if you compute error bars using one of the many methods that's included in the code, and your error bars are huge, that's a signal of misfit.
03 Feb 2012 Boosted Generalized Additive Models (bgam) package Boosting for the Generalized Additive and Linear Models (GAM and GLM). Author: Patrick Mineault

I replied to Brendan by email, but for other having the same problem: yes, you need to add back in the offset of the model (thefit.cs(end)) manually to do your predictions, since thefit.evaluate evaluates *only* the learners.
03 Feb 2012 Fit GLM with quadratic penalty Fits GLM with a quadratic penalty, determines hyperparams through cross-validation or evidence Author: Patrick Mineault

Sure, if you use X*results.w this will give you eta, the linear predictor. Then pass this through your inverse link, say the exponential if you use the Poisson distribution. That's your prediction.
Comments and Ratings on Patrick's Files View all
  • 1 - 5 of
  • 36
Updated File Comment by Comments Rating
22 Apr 2012 Fit GLM with quadratic penalty Fits GLM with a quadratic penalty, determines hyperparams through cross-validation or evidence Author: Patrick Mineault Mineault, Patrick

Either Generalized Linear Models by McCullough and Nelder or Generalized Additive Models: An introduction with R by Simon Wood.
22 Apr 2012 Fit GLM with quadratic penalty Fits GLM with a quadratic penalty, determines hyperparams through cross-validation or evidence Author: Patrick Mineault Marc

Hi Patrick, this is a great package!

Are you able to suggest a couple of references that are the basis for this penalized GLM approach and for your particular implementation?

Thanks.
22 Apr 2012 Fit GLM with quadratic penalty Fits GLM with a quadratic penalty, determines hyperparams through cross-validation or evidence Author: Patrick Mineault Marc
26 Mar 2012 Auto Gaussian & Gabor fits Fit 1D/2D Gaussian or Gabor to a curve/surface without start guesses for params Author: Patrick Mineault Mineault, Patrick

Ben, bootstrapping might work incorrectly if the Gaussian bump takes only a few pixels. In general, MCMC should be more reliable. The only reason I would use bootstrapping over MCMC is that MCMC is mathematically complicated and it's a bit tough to explain in the methods section of a paper.

Daniel, it's difficult to say without having access to the data you have. It's possible that there's a bug in the code that only appears for very specific datasets. Try to boil it down to the simplest example that shows the bug and send me the code to replicate the problem.
20 Mar 2012 Auto Gaussian & Gabor fits Fit 1D/2D Gaussian or Gabor to a curve/surface without start guesses for params Author: Patrick Mineault Daniel

Hi Patric, very nice program, however I think it is not acting "robustly" for me. It keeps conking out after about 10 iterations for each parameter, saying a local minimum is found, and saying that the iteration has yielded a tolerance less than my TolX... however I can't seem to create my own options structure with tolerances that this will use. The upshot is my R2 is negative and my errors are huge! Any thoughts?
Top Tags Applied by Patrick
  • 1 - 5 of
  • 75
generalized linear model, glm, regularization, 2d gaussian, argument validation
Files Tagged by Patrick View all
  • 1 - 5 of
  • 8
Updated   File Tags Downloads
(last 30 days) 		Comments Rating
07 Dec 2011 Screenshot Fit GLM with quadratic penalty Fits GLM with a quadratic penalty, determines hyperparams through cross-validation or evidence Author: Patrick Mineault generalized linear mo..., regularization, penalized regression, map, glm, linear regression 15 5
  • 5.0
5.0 | 2 ratings
02 Nov 2011 Screenshot Auto Gaussian & Gabor fits Fit 1D/2D Gaussian or Gabor to a curve/surface without start guesses for params Author: Patrick Mineault gaussian, curve fit, global optimization, gibbs sampling, surface fit, 2d gaussian 97 17
  • 4.8
4.8 | 5 ratings
22 Sep 2011 Screenshot Matching pursuit for 1D signals Performs Matching Pursuit on 1d (temporal) signals with custom basis Author: Patrick Mineault matching pursuit, basis pursuit, convolution, spectrogram, timefrequency analysi..., temporal signal 43 0
02 Sep 2011 Screenshot Fast B-spline class Fit, evaluate, differentiate non-uniform B-splines of any order - fast Author: Patrick Mineault bspline, pspline, spline, cubic spline, smoothing spline, interpolation 43 0
02 Jul 2011 Screenshot Boosted Generalized Additive Models (bgam) package Boosting for the Generalized Additive and Linear Models (GAM and GLM). Author: Patrick Mineault glm, gam, generalized linear mo..., generalized additive ..., boosting, maximum likelihood 33 4

Contact us at files@mathworks.com