Objective To predict neural tube birth defect （NTD） using support vector machine （SVM）. Method The dataset in the pilot area was divided into non overlaid training set and testing set. SVM was trained using the tr...Objective To predict neural tube birth defect （NTD） using support vector machine （SVM）. Method The dataset in the pilot area was divided into non overlaid training set and testing set. SVM was trained using the training set and the trained SVM was then used to predict the classification of NTD. Result NTD rate was predicted at village level in the pilot area. The accuracy of the prediction was 71.50% for the training dataset and 68.57% for the test dataset respectively. Conclusion Results from this study have shown that SVM is applicable to the prediction of NTD.展开更多
基金supported by CAS(KZCX2-YW-308)the MOST(2007DFC20180+2 种基金2007AA 12Z233)NSF(4047111170571076).
文摘Objective To predict neural tube birth defect （NTD） using support vector machine （SVM）. Method The dataset in the pilot area was divided into non overlaid training set and testing set. SVM was trained using the training set and the trained SVM was then used to predict the classification of NTD. Result NTD rate was predicted at village level in the pilot area. The accuracy of the prediction was 71.50% for the training dataset and 68.57% for the test dataset respectively. Conclusion Results from this study have shown that SVM is applicable to the prediction of NTD.
