Prediction of Ship Maintenance Cost by Support Vector Machines Based on Bayesian Optimization

Support vector machine is a widely used method in ship maintenance cost prediction. Its accuracy is affected by parameter setting and data quality. Commonly used parameter determination methods have the problems of long time or unstable results. In homogeneity often leads to unsatisfactory prediction results. In this paper, Bayesian optimization is used to determine the parameters of the support vector machine, which reduces the computational consumption and volatility. reduces the computational consumption and volatility; at the same time, k-means clustering is used to classify the original data, train and predict by class, which significantly improves the accuracy of the prediction model. By way of example verification, the prediction model proposed in this paper is By way of example verification, the prediction model proposed in this paper is compared with the existing model, and the practicability, scientificity and effectiveness of the method are verified.