How accurately can sitting and the intensity of walking and cycling be classified using an accelerometer on the waist for the purpose of the “Global recommendations on physical activity for health”?

摘要: The society has become more sedentary and developed a lack of physical activity, therefore increasing health risks. Feedback is needed to change these behaviours. For this feedback, first accurate monitoring needed: behaviour must be classified as well the intensity activity. In report State Art analysis performed compare different classification techniques finally two methods, both using an accelerometer on waist, are worked out. These methods Integral Modulus Accelerometer (IMA) machine learning technique (MLT): support vector (SVM). then applied in laboratory experiment study their quality. A measurement setup made create dataset following activities: standing, sitting, lying, walking (2.4 - 7.5 km/h) cycling (10.1-19 km/h). This (n=15) analysed Matlab for methods. The IMA method was unable monitor behaviour, but could classify activity (PA) with accuracy 66%. SVM within subjects able 91±20% PA 94±5%. between accuracies decrease 71±13% 45±35% classification. IMA implemented old feedback system, overall daily amount can significantly outperformed by replacing it current implementation. At moment however, only used improve cannot yet new additions such implementation specific intensities. New properties system have been found that might increase subject enable applicable extra options.