Perbandingan Algoritme Feature Selection Information Gain dan Symmetrical Uncertainty pada Data Ketahanan Pangan
Regional grouping based on indicators of food security is very important to take the proper policy in terms of deciding the targets and providing recommendations for tackling food insecurity. One method that can be used to classify objects into classes is decision tree algorithm. In this research, two decision tree models are constructed. The first decision tree utilizes information gain feature selection algorithm, whereas the second decision tree uses symmetrical uncertainty feature selection algorithm. These methods are used to classify the indicator of food security data from all districts in the provinces of Indonesia, obtained from the United Nations World Food Programme and Food Security Council. Then, the accuracy of both methods are compared. The result showed that the first decision tree is better than the second decision tree. The average accuracy of the first decision tree is 52.02%, while the average accuracy of second decision tree is 49.84%.
- UT - Computer Science