Abstract
The rapid industrialization of global livestock production has intensified the threat of viral epidemics, in which the intestinal, respiratory, and reproductive systems are susceptible to viral attacks. Understanding the mechanism of virus-host interactions will facilitate the development of prevention strategies against highly mutable and fast-spreading pathogens. This review examines recent progress in applying machine learning (ML) and deep learning (DL) to the study and control of animal viral diseases. By analyzing existing research, we show how these techniques improve the prediction of host-microbe interactions, support continuous monitoring of animal health, and accelerate the discovery of drug targets and vaccine candidates. Integrating ML and DL frameworks enables more accurate modeling of complex biological processes and offers new tools for data-driven veterinary science. Nevertheless, challenges remain, including unbalanced datasets, the structural and evolutionary complexity of viruses, and the poor cross-species transferability of predictive models. Future work should emphasize algorithmic designs suited to small-sample, multivariate time series data and promote the development of intelligent systems that unite virology, immunology, and epidemiology. The combination of reinforcement learning for optimizing vaccination strategies and unsupervised learning for detecting emerging pathogens may ultimately lead to adaptive, efficient, and precise systems for disease prevention, supporting both animal health and sustainable livestock development.