Multiple Linear Regression in Predicting Motor Assessment Scale of Stroke Patients

The Multiple Linear Regression (MLR) is a predictive model that was commonly used to predict the clinical score of stroke patients. However, performance slightly depends on method feature selection data as input predictor model. Therefore, appropriate needs be investigated in order give an optimum prediction. This paper aims (i) develop for Motor Assessment Scale (MAS) prediction patients, (ii) establish relationship between kinematic variables and MAS using model, (iii) evaluate based root mean squared error (RMSE) coefficient determination R2. Three types methods involve this study which are combination all variables, best four or less p < 0.05. MLR two assessment devices (iRest ReHAD) has been compared. As result, ReHAD with 0.05 Draw I (RMSEte = 1.9228, R2 0.8623), Diamond 2.6136, 0.7477), Circle 2.1756, 0.8268). These finding suggest stoke patients strong, able extracted from device.