Hayder Hussein Thary, Duraid Y. Mohammed & Khamis A. Zidan
Abstract
Flu and Typhoid Fever Infectious diseases are developed an ongoing concern for many health systems around the world due to the pressure of large visits to health centers by patients. This presents pressing needs for developing an adequate, quick, and accurate recognition technique to reduce overloads on health centers and doctors to diagnose the suspected patients using advanced computing technologies. This study firstly showed that Typhoid and flu are at the top of the common list of infectious diseases attacking the Iraqi community by utilizing a tailor-made questionnaire. Most of the previous studies have neglected infectious disease gravity that increased due to the lack and scarcity of data. In this study, the dataset was collected and processed as shown later, which was then employed as features in the diagnosis process. The obtained data are classified into three named classes: healthy people, patients with flu, and patients with Typhoid. A machine learning model is used for solving this diagnose problem by using the Support Vector Machine (SVM) technique. The proposed method used 16 features of each sample to classify the samples in a three-class named before. Hence, the feature selection algorithms were not included in this study due to the small size of the dimensionality of the features and to minimize the computed cost. The suggested method illustrates promising and excellent diagnosis performance with 94.1% of classification accuracy. Consequently, the technique proposed has shown precisely discriminates between Flu and Typhoid diseases.
أ د خميس عواد زيدان | Khamis A. Zidan | 9208