Comparison of Drug Type Classification Performance Using KNN Algorithm
| Autor: | Aldi, Febri, Nozomi, Irohito, Soeheri, Soeheri |
|---|---|
| Rok vydání: | 2022 |
| Předmět: | |
| Zdroj: | Sinkron : jurnal dan penelitian teknik informatika; Vol. 7 No. 3 (2022): Article Research Volume 7 Number 3, July 2022; 1028-1034 |
| ISSN: | 2541-2019 2541-044X |
| DOI: | 10.33395/sinkron.v7i3.11487 |
| Popis: | The error of decommissioning is a serious problem that is often faced in medicine. In the face of these problems, information technology has a very important role. One of the information technologies that can be used is to use the machine learning classification algorithm K-Nearest Neighbor KNN. KNN is a type of machine learning algorithm that can be applied to problems with classification and regression prediction. The classification of types of drugs for patients greatly affects the health of the patient. The patient data is processed and transformed to numbers, which are then divided into training data and test data from 90:10, 80:20, 70:30 and using the Cross Validation model. KNN works through the nearest neighboring value with a value of k = 3 calculated by the calculation of Euclidean Distance, and then evaluated using the Confusion Matrix. The performance of the KNN algorithm resulted in the highest Accuracy value of 98.33%, a Precision value of 98.8%, a Recall value of 96.2%, and an F-measure value of 97.48%. The performance is obtained from the sharing of training data and 90:10 test data. The data share results in high performance compared to other data shares, including using the Cross Validation model. And the lower the k value, the higher the value of the resulting performance. The results show that the performance of the KNN algorithm is working well. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|