With the continuous development and progress of society, the human resource data of enterprises is increasing. How to extract information from these human resource data has become an urgent need in the development and construction process of enterprises, and big data technology has brought new opportunities for predicting the demand for human resources in enterprises. Good human resource mining and prediction techniques can help enterprise managers have a clear understanding of the human resource situation of the enterprise, provide them with correct human resource demand prediction data, and help managers better manage the enterprise. This article applies a support vector machine (SVM) to establish a classification prediction model with the theme of predicting employee turnover probability. By visualizing the predicted results of the model, managers can understand the decision-making path of employee turnover and the ranking of important influencing factors, in order to provide targeted policy recommendations for enterprise human resource management. When using machine learning algorithms to predict and model practical problems, the quality of the model directly determines the accuracy and credibility of the prediction results. This article also uses the Analytic Hierarchy Process (AHP) to assign different weights to each indicator, and then it uses the Technique for Order Preference by Similarity to the Ideal Solution (TOPSIS) method to evaluate and sort the comprehensive performance of the model, selecting the algorithm model that performs best in the employee turnover problem dataset.