Читать книгу Data Analytics in Bioinformatics - Группа авторов - Страница 26

Decision Tree is a form of supervised machine learning and was invented by William Belson in the year 1959 [82]. It predicts the response values by learning the decision rules that were derived from features [83–84]. They are good for evaluating the options. It is used in operation research and decision analysis. An example of Decision Trees considering a person is having heart disease or not is presented below in Figure 1.15 for easy understanding.

The above figure depicts the answer to the Question “A person having Heart Disease or not?” by concerning various conditions and reaching a conclusion. Initially, it is checked that a person having chest pain or not. If yes, then it is checked that the person has high blood pressure or not. If the blood pressure if high or even low, then the person is suffering from heart disease. If the person doesn’t have chest pain then he is not suffering from heart disease. After implementing the Decision tree on the heart disease dataset [41] the AUC values are generated and presented in Table 1.6. The implementation was done in Python (Google Colab).

Figure 1.15 Decision tree.

Table 1.6 AUC: Decision trees.

Parameter	Data	Value	Result
The area under the ROC Curve (AUC)	Training Data	0.9588996	Outstanding
Test Data	0.9773333	Outstanding
	Index: 0.5: No Discriminant, 0.6–0.8: Can be considered accepted, 0.8–0.9: Excellent, >0.9: Outstanding

The obtained value of Training Data is 0.9588996 that attains an outstanding remark and the value of the testing data is 0.9773333 that attains an outstanding remark in the AUC score. The result indicates that the decision tree model performs outstandingly on the heart disease dataset.