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.