The application of Support Vector Machine for prediction of posttraumatic stress disorder on adults in flood district / 中华流行病学杂志

ABSTRACT

To predict the occurrence ofposttraumatic stress disorder (PTSD),using a Support Vector Machine (SVM) on adults in flood district.Diagnostic and Statistical Manuals on Mental Disorders (IV Edition) were used to examine and diagnose the victims in flood districts.Based on the forecasting model of SVM with PTSD as dependent variables and 23 influence factors of PTSD as independent variables,prediction of PTSD was conducted among the victims.After considering 23 influence factors into the prediction model,the agreement rate of prediction of the model was 88.05 percent,with sensitivity as 75.0 percent,and specificity as 89.4percent.Conclusion: The prediction model based on SVM with 23 influence factors had good effect on predicting the occurrence of PTSD.