Heart failure clinical Data Analysis
Download 428.37 Kb.
|
466 project
Classification ModelTo perform the classification, we have used two most renowned technique in classification domain which is Decision tree and support vector machine (SVM) [7] described below. Decision trees is one of the popular rule-based methods for classification and prediction. Decision tree modeling is a technique used widely in rule mining. The tree’ learned’ by splitting the source into subsets based on a value of attribute. each subtree learns from the subset recursively This process keeps repeating until subset only contains the data of a single class or further division does not improve the learning of a tree. Decision trees use statistical methods that facilitate the detection, categorization, and generalization of a data set. No subjective information or criteria are required for the creation of a decision tree, and are therefore it is sufficient for exploratory knowledge exploration. In fact, the decision trees classifier has good accuracy. The decision tree has achieved the accuracy score of 0.79. The following image shows the decision tree of our dataset containing 13 attributes including the class attribute. install.packages('C50') library('C50') # load the package data = read.csv('heart_failure_clinical_records_dataset.csv',header = TRUE) dTree <- C5.0(data[,-13], as.factor(wine[,13])) # train the tree summary(dTree) # view the model components plot(dTree,tp_args=list(fill=c("Light Pink","Light Blue")) , main = 'heart failure decision tree') # view the model graphically Figure 17: Decision Tree Download 428.37 Kb. Share with your friends:
|