Human immunodeficiency virus (HIV) is one of the fastest developing pathogens known. HIV/AIDS is an incurable disease which causes severe damage to the immune system. The recommended treatment for HIV/AIDS is a combination of three antiretroviral (ARV) drugs from two or more different drug groups and is known as highly active antiretroviral therapy (HAART). Drug resistance is a major impediment experienced by therapist in treating HIV infected patients. Theoretically, drug resistance can be predicted from the presence of specific mutations in the viral genome. With the current disease burden and lack of resources in developing countries, phenotypic tests are not viable. Developing a computational prediction of drug resistance phenotype will enable efficient and timely selection of the best treatment regimens. Nevertheless, the very large range of possible drug combinations and of viral mutational patterns leads to an extremely complex scenario, making prediction of in vivo treatment response extremely challenging. To deal with such complexity, machine learning methods are being increasingly explored. Clinical and technological developments has generated and stored large volumes of data in public databases which facilitates the use of machine learning methods for predicting drug resistance. Quite a lot of machine learning approaches such as neural networks, support vector machine, Bayesian networks, decision trees and linear regression have been proposed for the prediction of phenotypic drug resistance. Therefore, conducting resistance testing is certainly significant in order to administer appropriate antiviral drugs to HIV infected patients. © Springer Nature Switzerland AG 2019. All rights reserved.