Autoencoder networks for HIV classification

Brain Leke Betechuoh, Tshilidzi Marwala, Thando Tettey

Research output: Contribution to journalArticlepeer-review

56 Citations (Scopus)


In this paper, we introduce a new method to analyse HIV using a combination of autoencoder networks and genetic algorithms. The proposed method is tested on a set of demographic properties of individuals obtained from the South African antenatal survey. When compared to conventional feedforward neural networks, the autoencoder network classifier model proposed yields an accuracy of 92%, compared to an accuracy of 84% obtained from the conventional feedforward neural network models. The area under the ROC curve for the proposed autoencoder network model is 0.86 compared to an area under the curve of 0.8 for the conventional feedforward neural network model. The autoencoder network model for HIV classification, proposed in this paper, thus outperforms the conventional feedforward neural network models and is a much better classifier.

Original languageEnglish
Pages (from-to)1467-1473
Number of pages7
JournalCurrent Science
Issue number11
Publication statusPublished - 12 Oct 2006
Externally publishedYes


  • Autoencoder networks
  • Genetic algorithms
  • HIV classification

ASJC Scopus subject areas

  • Multidisciplinary


Dive into the research topics of 'Autoencoder networks for HIV classification'. Together they form a unique fingerprint.

Cite this