基于PSO-FWSVM的糖尿病预测模型

Diabetes Prediction Model Based on PSO-FWSVM

糖尿病是三大慢性病之一,及早发现有利于对该病进行控制。为了提高早期诊断率,提出基于支持向量机(SVM)建立合适的糖尿病预测模型。在分析糖尿病数据特点基础上,提出对核函数进行加权处理,避免弱相关特征对分类结果的影响,从而提高了分类模型的识别率。之后采用自适应粒子群优化算法对FWSVM模型进行参数优化,并对某医院的实际糖尿病数据集进行模型的训练和识别。实验结果表明:相对于其他常用模型,论文的模型识别准确率和运算效率都有一定程度的提高,达到90.36%,性能上优于其他几种模型。

Diabetes is one of the three major chronic diseases,and early detection is conducive to the control of the disease.In order to improve the early diagnosis rate,a suitable diabetes prediction model based on support vector machine(SVM)is proposed.Based on the analysis of the characteristics of diabetes data,the kernel function is weighted to avoid the influence of trivial relevant features on the classification,thus the recognition rate of the classification model is improved.Later,the adaptive particle swarm optimization algorithm is used to optimize the parameters of the FWSVM model.The proposed model is trained and identified based on the empirical diabetes dataset of a hospital.The results show that compared with other common models,the recognition accuracy and computational efficiency of the proposed model are improved,it is reaching 90.36%,and the performance is better than other models.