智慧城市面临越来越严峻的信息安全威胁,传统的信息安全风险预测方法较难适应智慧城市信息安全的复杂性、非线性与不确定性等特性。本文通过改进现有智慧城市信息安全风险指标体系,基于机器学习中的随机森林算法,构建一个智慧城市信息安全风险预测模型,并与传统模型预测结果对比,结果表明,模型预测效果较其他模型更优秀。 Smart cities are facing more and more serious threats to information security. The traditional information security risk prediction methods are more difficult to adapt to the complexity, nonlinearity and uncertainty of smart city information security. This paper builds an information security risk prediction model of smart city based on random forest algorithm in machine learning by improving the existing information security risk index system of smart city and compares it with the traditional model prediction results. The results show that the model prediction effect is more effective than other models excellent.