Minyatür eylemsizlik duyucuları ve manyetometreler ile insan aktivitelerinin sınıflandırılması üzerine karşılaştırmalı araştırma

This study provides a comparative assessment on the different techniques ofclassifying human activities that are performed using body-worn miniature inertialand magnetic sensors. The classification techniques compared in thisstudy are: naive Bayesian (NB) classifier, artificial neural networks (ANNs),dissimilarity-based classifier (DBC), various decision-tree methods, Gaussianmixture model (GMM), and support vector machines (SVM). The algorithms forthese techniques are provided on two commonly used open source environments:Waikato environment for knowledge analysis (WEKA), a Java-based software;and pattern recognition toolbox (PRTools), a MATLAB toolbox. Human activitiesare classified using five sensor units worn on the chest, the arms, and the legs.Each sensor unit comprises a tri-axial gyroscope, a tri-axial accelerometer, and atri-axial magnetometer. A feature set extracted from the raw sensor data usingprincipal component analysis (PCA) is used in the classification process. Threedifferent cross-validation techniques are employed to validate the classifiers. Aperformance comparison of the classification techniques is provided in terms oftheir correct differentiation rates, confusion matrices, and computational cost.The methods that result in the highest correct differentiation rates are found tobe ANN (99.2%), SVM (99.2%), and GMM (99.1%). The magnetometer is thebest type of sensor to be used in classification whereas gyroscope is the leastuseful. Considering the locations of the sensor units on body, the sensors wornon the legs seem to provide the most valuable information.Keywords: inertial sensors, gyroscope, accelerometer, magnetometer, activityrecognition and classification, feature extraction and reduction, cross validation,Bayesian decision making, artificial neural networks, support vector machines,decision trees, dissimilarity-based classifier, Gaussian mixture model, WEKA,PRTools.