Machine learning algorithms have gained popularity in recent years in many fields due to their promising results in predictive performance of classification problems. The application of machine-learning algorithms has also been highly simplified in the last years due to their well-documented integration in commonly used statistical programming languages (such as R or Python). Machine learning is a subsection of Artificial Intelligence (AI), it is one of the most promising tools in classification and it a model that aims to discover the unknown function, dependence, or structure between input and output variables. This study proposes statistical and machine learning models to diagnose anemia disease. Some machine learning techniques have been used in this work to avoid overfitting, pre-process the data and adjust the outliers to give better results. Three classifiers, including Logistic Regression, k-Nearest Neighbor and Decision Tree are implemented in this work. The performance of the models is evaluated based confusion matrix, recall, precision, f1-score, accuracy, Matthews correlation coefficient and ROC curve to compute area under the curve (AUC). The results show the logistic regression has the highest accuracy of 99.57%, with recall values of 99.41%, precision values of 99.61, f1-score of 99.51% and Matthews correlation coefficient values of 99.13%. Decision tree has the second highest accuracy of 98.64%, with recall values of 99.02%, precision values of 97.87, f1-score of 98.44%, and Matthews correlation coefficient values of 97.23%. |
