Title: Predicting generalised anxiety disorder among women using decision tree-based classification

Authors: Neesha Jothi; Wahidah Husain; Nur'Aini Abdul Rashid; Lee Ker Xin

Addresses: School of Computer Sciences, University Sains Malaysia, Penang, Malaysia ' School of Computer Sciences, University Sains Malaysia, Penang, Malaysia ' College of Computer and Information System, Department of Computer Sciences, Princess Noura bint Abdulrahman University, Riyadh, Kingdom of Saudi Arabia ' School of Computer Sciences, University Sains Malaysia, Penang, Malaysia

Abstract: Mental health presents as one of the greatest challenges to the current generation. It has been reported that about 5% of the population in developed countries are affected by generalised anxiety disorder (GAD) with women twice as likely to be affected as compared to men. Predicting GAD among women is no longer an arduous task especially with the assistance of data mining technology. In this paper, a methodology encompassing data collection, data pre-processing, data analysis and data mining process using random forest approach is drawn for an effective prediction. The random forest approach is one of the classification data mining techniques which is embedded with good predictive characteristic. The result of this study in term of accuracy, sensitivity and specificity conforming to its high predictive performance in GAD prediction based on depressive symptoms. Besides that, several popular machine learning techniques are also applied to the resultant dataset of this study and the comparison result attests to random forest algorithm outperformed other methods. The generated prediction model is expected to provide an effective screening process to detect generalised anxiety disorder earlier among women in Malaysia.

Keywords: data mining; data mining in healthcare; generalised anxiety disorder; GAD; random forest.

DOI: 10.1504/IJBIS.2018.093998

International Journal of Business Information Systems, 2018 Vol.29 No.1, pp.75 - 91

Received: 24 Oct 2016
Accepted: 10 Feb 2017
Published online: 13 Aug 2018 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article