خط مشی دسترسیدرباره ما
ثبت نامثبت نام
راهنماراهنما
فارسی
ورودورود
صفحه اصلیصفحه اصلی
جستجوی مدارک
تمام متن
منابع دیجیتالی
رکورد قبلیرکورد بعدی
Document Type : Latin Dissertation
Language of Document : English
Record Number : 152617
Doc. No : ET24409
Main Entry : ROZITA D A M
Title Proper : THE USE OF UNLABELLED DATA FOR SUPERVISED LEARNINGPOMPES PAR TRANSITIONS MULTIPLES
Note : This document is digital این مدرک بصورت الکترونیکی می باشد
Abstract : When provided with enough labelled training examples, a supervised learning algorithmcan learn reasonably accurately. However, creating sufEcient labelled data to train accurateclassifiers is time consuming and expensive. On the other hand, unlabelled data is usuallyeasy to obtain. This research introduces a novel approach, Guelph Cluster Class (GCC),which improves the task of classification with the use of unlabelled data- The novelty ofthis approach Lies in the use of an unsupervised network, Self-Organizing Map, to selectnatural clusters in labelled and unlabelled data. Subclasses (made by labelled data) areused to assign labels to udabelled patterns to produce self-labelled data. The performanceof several variants of the GCC system have been obtained by running a Back-Propagationnetwork on labelled and self-labelled data. Results of experiments on several benchmarkdatasets demonstrate an increasing power for the classification procedure even when thenumber of labelled data is very small..
Subject : Electericl tess
: برق
electronic file name : TL48616.pdf
Title and statement of responsibility and : THE USE OF UNLABELLED DATA FOR SUPERVISED LEARNINGPOMPES PAR TRANSITIONS MULTIPLES [Thesis]
 
 
 
(در صورت عدم وضوح تصویر اینجا را کلیک نمایید)