Home

About

Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network
FAQHELPSITE MAPCONTACT US


  Advanced Search  

 
Updating Kernel Methods in Spectral Decomposition by Affinity Perturbations
 

Summary: Updating Kernel Methods in Spectral Decomposition
by Affinity Perturbations
Yaniv Shmuelia,
, Guy Wolfa
, Amir Averbucha
aSchool of Computer Science, Tel Aviv University, Tel Aviv 69978, Israel
Abstract
Many machine learning based algorithms contain a training step that is done
once. The training step is usually computational expensive since it involves
processing of huge matrices. If the training profile is extracted from an evolving
dynamic dataset, it has to be updated as some features of the training dataset
are changed. This paper proposes a solution how to update this profile effi-
ciently. Therefore, we investigate the problem of updating the training profile
when the data is constantly evolving. The data is modeled by kernel method
and processed be spectral decomposition. In many algorithms for clustering and
classification, a low dimensional representation of the affinity (kernel) graph of
the embedded training dataset is computed. Then, it is used to classify newly
arrived data points. We present methods for updating such embeddings of the
training datasets in an incremental way without the need to perform the entire
computation upon changes in small number of the training samples. Efficient

  

Source: Averbuch, Amir - School of Computer Science, Tel Aviv University

 

Collections: Computer Technologies and Information Sciences