Title: Data mining techniques for vestibular data classification

Authors: Domenico Mirarchi; Giovanni Canino; Patrizia Vizza; Pierangelo Veltri; Salvatore Cuomo; Claudio Petrolo; Giuseppe Chiarella

Addresses: Department of Medical and Surgical Sciences, University of Catanzaro, Viale Europa, 88100 Catanzaro, Italy ' Department of Medical and Surgical Sciences, University of Catanzaro, Viale Europa, 88100 Catanzaro, Italy ' Department of Medical and Surgical Sciences, University of Catanzaro, Viale Europa, 88100 Catanzaro, Italy ' Department of Medical and Surgical Sciences, University of Catanzaro, Viale Europa, 88100 Catanzaro, Italy ' Department of Mathematics and Applications, University of Naples Federico II, Via Cintia, 80126 Napoli, Italy ' U.O. Audiology and Phoniatrics, University of Catanzaro, Viale Europa, 88100 Catanzaro, Italy ' U.O. Audiology and Phoniatrics, University of Catanzaro, Viale Europa, 88100 Catanzaro, Italy

Abstract: The vestibular system controls motor functions to providing the sense of balance and spatial orientation. The damages of this system are generally estimated with the vestibular evoked myogenic potentials (VEMPs) test. In the proposed work the data of this test are analysed using a specific algorithm and developing a dedicate framework to support clinical activity, with an easy-to-use GUI. The data have been collected and used for extracting a prediction of a pathology distribution by using data mining techniques. In to the studies two different algorithms. In a first step, the Bayesian classifier has been chosen to extract the predisposition of female to lodge diseases of the auditory system compared to the male. In the final contribution, the nearest neighbour classification algorithm has been choosing for as classifier. Results show a better classification of the pathologies as a function of sex respect to the results obtained with the previous classifier.

Keywords: data mining; vestibular evoked myogenic potentials; VEMPs; disease classification; Bayesian method; k-Nearest Neighbour algorithm.

DOI: 10.1504/IJITST.2017.085734

International Journal of Internet Technology and Secured Transactions, 2017 Vol.7 No.1, pp.51 - 70

Received: 10 Nov 2016
Accepted: 14 Feb 2017

Published online: 10 Aug 2017 *

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