Detection of Canine Parvovirus using ELM and Support Vector Machine Algorithm from Blood Tests

Canine parvovirus is a widespread disease affecting dogs, particularly young puppies. Most frequently, a fecal ELISA test is used to detect this virus. This test does not guarantee the presence of the virus in the dog and costs between 3500 and 4000 RS (INR). Generally, Veterinary doctors conduct blood tests of the affected dogs and then recommend the fecal ELISA test since different parameters of blood tests also cannot accurately predict the presence of Parvovirus. In this paper, a novel method of Parvovirus detection is proposed based on the blood chemistry report of the dog. This method uses a Semi-supervised Support Vector Machine (SSVM) based model to detect the virus using the blood chemistry parameters of the animal with labeled and unlabeled data.