Investigation into the use of autoencoder neural networks, principal component analysis and support

Vukosi Ntsakisi Marivate, Fulufhelo Vincent Nelwamondo, Tshilidzi Marwala

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Data collection often results in records that have missing values or variables. This investigation compares 3 different data imputation models and identifies their merits by using accuracy measures. Autoencoder Neural Networks, Principal component analysis and Support Vector regression are used for prediction and combined with a genetic algorithm to then impute missing variables. The use of PCA improves the overall performance of the autoencoder network while the use of support vector regression shows promising potential for future investigation. Accuracies of up to 97.4 % on imputation of some of the variables were achieved.

Original languageEnglish
Title of host publicationProceedings of the 17th World Congress, International Federation of Automatic Control, IFAC
Edition1 PART 1
DOIs
Publication statusPublished - 2008
Event17th World Congress, International Federation of Automatic Control, IFAC - Seoul, Korea, Republic of
Duration: 6 Jul 200811 Jul 2008

Publication series

NameIFAC Proceedings Volumes (IFAC-PapersOnline)
Number1 PART 1
Volume17
ISSN (Print)1474-6670

Conference

Conference17th World Congress, International Federation of Automatic Control, IFAC
Country/TerritoryKorea, Republic of
CitySeoul
Period6/07/0811/07/08

Keywords

  • Neural networks in social systems

ASJC Scopus subject areas

  • Control and Systems Engineering

Fingerprint

Dive into the research topics of 'Investigation into the use of autoencoder neural networks, principal component analysis and support'. Together they form a unique fingerprint.

Cite this