混沌理论和支持向量机的就业率预测

Research on Employment Forecast Based on Chaos Theory and Support Vector Machine

就业率预测具有较强复杂性,为此提出了基于混沌理论和支持向量机的就业率预测方法,用以降低就业率预测误差。按照混沌变量运动自身规律使用越界处理优化混沌算法跳出局部最优提升搜索精度;运用支持向量机拟合就业率预测非线性关系,构建基于支持向量机的就业率预测模型,并使用K邻近算法,构建样本数据集;运用经过优化的混沌粒子群算法优化支持向量机参数,训练数据样本集,构建改进支持向量机的就业率预测模型实现预测。通过实验验证,该方法具有较强的训练能力和较低的模型复杂程度且训练时间较快,预测高校毕业生就业率误差较低,具有良好的预测精度。

Employment rate prediction has strong complexity, this paper applies chaos theory and support vector machine-based employment rate prediction methods to reduce the employment rate prediction error. By using the algorithm to fit the nonlinear relationship, we construct the employment rate prediction model based on SCM, use the k neighboring algorithm, optimize the parameters, train the data sample set and build the employment rate prediction model to achieve prediction. Through experimental verification, the proposed method has strong training ability and low model complexity and fast training time, low prediction employment rate error and good prediction accuracy.