Abstract: The problem of bacterial resistance poses a significant threat to human and animal health as well as public safety, making the discovery of effective new antimicrobial compounds an urgent priority. Traditional methods for screening antimicrobial activity are often time-consuming and labor-intensive, with limited accuracy and objectivity. As a branch of artificial intelligence, machine learning algorithms have demonstrated exceptional capabilities in processing large-scale data, feature extraction, and model optimization, leading to their increasing application in the screening of antimicrobial substances. This paper reviews commonly used machine learning models, such as random forests, support vector machines, and deep learning, in antimicrobial activity screening. It provides an in-depth exploration of machine learning applications in the discovery of antimicrobial peptides, essential oils, and polyphenols, aiming to offer valuable insights into the application of machine learning techniques for identifying antimicrobial compounds.

Key words: machine learning; antimicrobial substances; screening