Alpha Thalassemia Classification Based on Rough Set Theory

This paper proposed a classification model of alpha thalassemia based on rough set theory. The model is start by applying preprocessing against noise and missing data. Then, rough set and Recursive Feature Eliminator (REF) are applied for feature selections. It is followed by predicting the existence of alpha thalassemia based on Random Forest algorithm. A collected dataset from Kaggle with alpha thalassemia features has been used in the proposed model. Experimental results show that Random Forest algorithm compared against machine learning algorithms before applying the feature selection and after applying it produced the highest accuracy of 96.36%, precision of 0.9809, F1-score of 0.9716, and specificity of 0.9655. The proposed model confirms that patient with alpha thalassemia can be easily predicted. The high F1-score obtained suggests that it is possible to predict the alpha thalassemia carrier using machine learning models.