Please use this identifier to cite or link to this item: https://hdl.handle.net/10356/46776
Title: Impacts of noise on machine learning
Authors: Qiao, Kaimu
Keywords: DRNTU::Engineering::Electrical and electronic engineering::Electronic systems::Signal processing
Issue Date: 2010
Abstract: Noise in real-world data is quite common and may impact the interpretations of the data, classifiers constructed from the data and classification decisions made based on the data. It can decrease the classifier's performance in many aspects such as classification accuracy, time in building a classifier and so on. Handling noise has become a key issue in machine learning and data mining. Many learning algorithms have been modified to enhance their performance in the environment of noise, like robust learners. However, when the noise level goes high, the performance of these algorithms may not be as good as expected. Thus, a better way is to preprocess the data before a classifier is constructed. Conventional mechanisms include noise filter and data polishing.
Description: 96 p.
URI: http://hdl.handle.net/10356/46776
Rights: Nanyang Technological University
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:EEE Theses

Files in This Item:
File Description SizeFormat 
EEE_THESES_125.pdf
  Restricted Access
11.19 MBAdobe PDFView/Open

Page view(s) 20

221
checked on Oct 23, 2020

Download(s) 20

13
checked on Oct 23, 2020

Google ScholarTM

Check

Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.