The Predictive Model of Esophageal Squamous Cell Carcinoma Differentiation

The diagnosis of the degree of differentiation of tumor cells can help physicians to make timely detection and take appropriate treatment for the patient’s condition. The original datasets are clustered into two independent types by the Kohonen clustering algorithm. One type is used as the development sets to find correlation indicators and establish predictive models, while the other type is used as the validation sets to test the correlation indicators and model. Thirteen indicators significantly associated with the degree of differentiation of esophageal squamous cell carcinoma are found by Kohonen algorithm in the development sets. Ten different machine learning classification algorithms are used to predict the differentiation of esophageal squamous cell carcinoma. The artificial bee colony-support vector machine (ABC-SVM) has better prediction accuracy than the other nine algorithms and has a shorter training time. The average accuracy of the 10-fold cross-validation reached 81.5\(\%\) by ABC-SVM algorithm. In the development sets, a model with the great merit for the degree of differentiation is found based on logistic regression algorithm. The AUC value of the model is 0.672 and 0.753 in the development sets and validation sets, respectively. \(p-values\) are less than 0.05. The results are shown that the model has a high predictive value for the differentiation of esophageal squamous cell carcinoma.