Участник:Tatarchuk

Материал из MachineLearning.

(Различия между версиями)
Перейти к: навигация, поиск
(Overview)
(Содержимое страницы заменено на «Tatarchuk»)
 
Строка 1: Строка 1:
[[Категория:Страницы участников|Tatarchuk]]
[[Категория:Страницы участников|Tatarchuk]]
-
'''Татарчук Александр Игоревич'''
 
-
 
-
= Selective Support Vector Machines =
 
-
=== Description ===
 
-
'''Authors'''
 
-
[mailto:vmottl@ya.ru V. Mottl]
 
-
[mailto:aitech@ya.ru A. Tatarchuk]
 
-
[mailto:andreyel@gmail.com A. Eliseev]
 
-
 
-
'''Developed at'''
 
-
[http://www.ccas.ru Computing Center of Russian Academy of Sceince], Moscow, Russia
 
-
 
-
'''Version''' 1.0
 
-
 
-
'''Date''' 14.10.2008
 
-
 
-
=== Overview ===
 
-
 
-
SelSVMs are modifications of classical Support Vector Machine provided with controlled selectivity property of feature/kernel selection.
 
-
 
-
The main features of the algorithms are the following:
 
-
 
-
* work on small training sets with large number of features/kernels;
 
-
* provide feature selection with controlled selectivity level;
 
-
* use Matlab optimization solver and Mosek solver;
 
-
* support precompiled kernels;
 
-
* return a posteriori probability estimates;
 
-
 
-
There are four algorithms in SelSVM library:
 
-
 
-
* Selective Relevance Feature Machine (SelRFM) -
 
-
* Selective Relevance Feature Machine (SelRKM) -
 
-
* Selective Support Feature Machine (SelSFM) -
 
-
* Selective Support Kernel Machine (SelSKM) -
 
-
 
-
=== How to use ===
 
-
 
-
 
-
=== Matlab code ===
 
-
 
-
 
-
=== Publications ===
 

Текущая версия

Личные инструменты