Gastric cancer is one of the most common malignant tumors of digestive tract in the world. With the development of science and technology，a large number of gastric cancer biomarkers have been excavated and applied to clinical practice. Recently，machine learning has been widely used in the research of gastric cancer biomarkers because of its efficient feature discovery and learning reasoning ability. This review summarizes the characteristics of commonly used machine learning algorithms in biomarker mining，analyzes the application value of biomarker mining in the diagnosis，efficacy monitoring and prognosis of gastric cancer and estimates the future research direction.