Development of machine learning Classifier models for classification Of motion sickness levels using Biosignals

Abstract

Motion sickness is one of the common sicknesses observed in newlinenumerous people who are in travel, driving, and even a fast walk, leading to newlinehave a discomfort and coming across vomiting, nausea and other newlinepsychological disorders. It is always required to overcome the motion newlinesickness across the individuals and on identifying the same, treatment has to newlinebe initiated to protect them overcome the observed discomforts. Also, the newlinelevel of motion sickness that is present in the individuals is also required since newlineonly based on the motion sickness levels treatment will be provided. This newlineresearch thesis has developed novel classifier models based on clustering newlinemechanism, neural computing and hybrid techniques for performing newlineclassification of the motion sickness levels and thereby help the individual to newlineovercome the sickness rate with necessary treatment being initiated. newlineThe research contribution focused in this thesis chapter is to design newlineand develop novel classifier models for classification of the motion sickness newlinelevels based on the multiple biosignals pertaining to the candidates EEG newlinesignal, centre of pressure and head and waist trajectory movement. Each newlinecandidate possesses multiple biosignal feature vectors and these signals are newlinepresented as input to the classifier models and the set performance metrics are newlineevaluated. This thesis was intended to develop effective classifier models and newlineit modelled new classifiers achieving better accuracy rate with minimized newlinemean square error value. The data samples presented to the modelled new newlineclassifiers were generated for 20 candidates and 90 candidates and were newlineemployed for testing. newline