Article Text
Abstract
Background and aims: Asthma is a chronic lung disease and has a raising worldwide prevalence. Lack of timely and appropriate control for this condition leads to financial and physical injuries. The aim of this study is to prediction of asthma control levels by applying data mining algorithms.
Methods: This is a cross-sectional study carried out in the city of Sanandaj in Iran. Samples consist of 600 referred patient patients who live with asthma to Tohid pulmonary clinic in Sanandaj In a period of two months in 2015. Data were collected based on the study's inclusion criteria. Preprocessing was performed and various algorithms include Support Vector Machine (SVM), Decision Tree (DT), K-Nearest Neighbor (KNN) and Naïve Bayesian was assessed. Finally results were evaluated by confusion matrix.
Results: Features ranked by applying feature selection methods; after in next step, 19 Features of 24 was chosen as the most effective asthma control features. Cough has the highest InfoGain, Relief-F and GainRatio comparing with other features. Results shows KNN and NaiveBayes have the highest accuracy near to 98%.
Discussion: Experts can analysis and design accurate decision support systems by using data mining methods in healthcare. These methods aims to reduction and optimal usage of data. Important factors in determination of asthma control level were identified by considering of accurate mining algorithms. Therefore, identification of high risk patients were performed and proper services were provided to them to prevent major complications.
- Data mining
- Asthma
- Feature selection.
This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/