Comments for Mathieu's log

Comment on Regularized Least Squares by Mohsen

Mohsen — Tue, 02 Oct 2012 22:30:59 +0000

Was a great post. I am struggling with kernel ridge for a while. I have a question regarding the regularizer. As far as I have read there two method for regularization of kernel matrix. One is , adding a small value to the diagonal of kernel matrix.and the second one is using truncated eigendecomposition which means just to put out the small eigenvalues .But I kind of stuck in the second case.I actually can’t understand why it’s a case.I know it kind of related to matrix inversion.Do you have any idea about that?

Comment on LSA and pLSA in Python by aeweiwi

aeweiwi — Wed, 18 Jul 2012 08:17:46 +0000

Hi,
I have been looking at your pLSA code in python, and debugging your program as some has noticed that there exist some failure in the folding-in process for new sample. I think I have resolved it by noticing that your normalize function have a strange behavior when you ask for normalizing an array of shape (1,Z) in the query. I noticed it set all values almost to zero except one (Thats why the maximization step sometimes goes below zero and sometimes have only one iteration to converge-which is strange), which is strange for a random initializer. Probably you may like to fix this error some day as so many guys ( including me ) are frequently reading your post.

Comment on Transparent system-wide proxy by Lucy

Lucy — Sun, 10 Jun 2012 07:33:17 +0000

I will tell you something: why don’t make a Windows software (GNU Freeware) that does exactly this function. You Linux geeks are genius, but helping people it’s more than type some kind of linux “command”. There are Windows people that need this too. We Windows users need help too. Yeah, I’m newbie, Linux may be great, but there are too many complex “command line” instructions to me (I know the graphical interface is as easy as Windows), but get complex things working in Linux is a pain. In Windows someone program an application, and it’s done. I know I may sound lazy, but not everyone is a genius. This is only an example. There are a lot of functions in Windows that need an application to get it work. For example, I was looking for a transparent proxy to VMware o Virtual PC, that works at ethernet level. In linux it may be easy for a genius, but in Windows it’s impossible without an application or a complex configuration. If you know a way to do that in Windows, it will be great!. Thank you…….. :)

Comment on LSA and pLSA in Python by PLSA and PLSM | Recognizing Attack Patterns: Clustering of Optical Flow Vectors in RoboCup Soccer

Tue, 05 Jun 2012 15:45:01 +0000

[...] solution to this is the use of PLSA combined with connected component analysis. Thanks to M. Blondel, I now have an implementation of the PLSA algorithm. To use it, I will have to map the current [...]

Comment on Dictzip reader in Ruby by Mathieu

Mathieu — Thu, 12 Apr 2012 15:54:35 +0000

No, what I wrote is correct. How would you read the compressed chunk if the size was the one of the uncompressed form?…

Comment on Dictzip reader in Ruby by coder

coder — Thu, 12 Apr 2012 13:56:53 +0000

There is a significant typo in the 3rd paragraph
(in its compressed form)
should read
(in its UNcompressed form)

Comment on Git memo by Mathieu

Mathieu — Mon, 02 Jan 2012 10:53:06 +0000

Valera: Sure no prob!

I fixed the escaping problems.

Comment on Git memo by Valera

Valera — Sun, 01 Jan 2012 10:15:53 +0000

Good memo. Could I translate it to Russian and put it on my site for my students (with reference, of course) ?

Commands need some escaping, since — is dispalyed as long dash and “” are tranlated to literary quotes.

Comment on Introduction to Dynamic Time Warping by noname

noname — Sun, 11 Dec 2011 16:56:02 +0000

Hi, an open-source implementation of various dtw algorithms is available for R at http://dtw.r-forge.r-project.org . It has an extensive manual and is callable from Python. Should work well for rapid experimentation.

Comment on Semi-supervised Naive Bayes in Python by Purvi

Purvi — Sat, 26 Nov 2011 16:27:25 +0000

Hello rick wei,

As you suggest, I have provided input data for td and delta in train function, but I am getting error of objects are not aligned. Here is output of that

File “/var/test/seminb.py”, line 156, in train
self.p_w_c[w,c] += td[w,d] * delta[d,c]
File “/usr/lib/python2.6/dist-packages/numpy/core/defmatrix.py”, line 290, in __mul__
return N.dot(self, asmatrix(other))
ValueError: objects are not aligned

I am running it on ubuntu linux system.

Please advise what should i do to successfully run train and other functions.