Latex fonts for MS Office

If you have struggled with making your MS PowerPoint or Word document
look visually compatible with the fonts used for math in LaTex - you
might want to get the AMS set of LaTex fonts for MS-Office

http://www.ams.org/index/tex/amsfonts.html

Peer review ? Really ?

Found in a "peer-reviewed" journal Advances in Neural Computing – italics mine.

 

Functional magnetic resonance imaging (fMRI) is a new non-hurt measure technique  for  brain  activity ， which  have  been  used  at  the  study  of  brain  cognition,  locating  nerve activity, medicine, psychology and other domains, and has become one of the  most important way of the study of brain function

…

 

In general, the general linear model take Gamma function as hemodynamic response function  to  get  design  matrix.  But,  it  is  not  reasonable  in  most  actually  case.  The BOLD signal of the cerebral activation is a collective response of an activated region and it can be explained as a mutual interaction process between the neural response to  a  stimulus  and  the  hemodynamic  change  due  to  the  activation  of  a  neural  cluster.

 

 

It goes in this spirit … but need I bother ?

Self tuning spectral clustering

"Self tuning spectral clustering" is a paper in NIPS2004 by L. Zelnik-Manor and P. Perona that realizes the very intuitive idea of multi-scale (scale space) clustering:

Abstract
Spectral clustering has been theoretically analyzed and empirically proven useful. There are still open issues:
(i) Selecting the appropriate scale of analysis,
(ii) Handling multi-scale data,
(iii) Clustering with irregular background clutter, and,
(iv) Finding automatically the number of groups.
We explore and address all the above issues. We first propose that a `local' scale should be used to compute the affinity between each pair of points. This local scaling leads to better clustering especially when the data includes multiple scales and when the clusters are placed within a cluttered background. We further suggest exploiting the structure of the eigenvectors to infer automatically the number of groups. This leads to a new algorithm in which the final randomly initialized k-means stage is eliminated.