Generating Individual Gait Kinetic Patterns Using Machine Learning C´ esarBou¸cas 1(B ) , Jo˜ ao P. Ferreira 1,2 , A. Paulo Coimbra 1 , Manuel M. Cris´ ostomo 1 , and Paulo A. S. Mendes 1 1 Institute of Systems and Robotics, Department of Electrical and Computer Engineering, University of Coimbra, Coimbra, Portugal cesar.boucas@isr.uc.pt 2 Department of Electrical Engineering, Superior Institute of Engineering of Coimbra, Coimbra, Portugal ferreira@isec.pt Abstract. In this study, data of 42 healthy individuals walking over a treadmill was used to train and test a neural network that produced individual kinetic patterns of gait cycle as output for a set of atomic features (gender, age, mass, height and gait speed) used as input. The proposed method implements a 3-layer feedforward architecture capable to produce the 3D gait patterns of ankle, knee and hip moment at once, with an average root mean squared error (RMSE) of 7% and average cor- relation coefficient (ρ) of 0.94 with respect to the ground truth patterns of the test set. The presented strategy may be used to support individual gait clinical analysis as an alternative to the use of the normal literature pattern that do not take into account the specific characteristics of the patients. Keywords: Human gait kinetics · Time series generation · Machine Learning 1 Introduction The study of human gait covers the way people use the movement of limbs to perform terrestrial locomotion. Gait analysis is one of the most active research fields in Biomechanics and have a broad range of applications, such as pathology detection [14], rehabilitation [16], prosthesis design, biometric identification [3] and bipedal robotic locomotion. Clinical gait analysis methods aims to provide an objective record that quan- tifies the magnitude of patients deviations from normal gait. Whereas a set of TheFunda¸c˜ ao para a Ciˆ encia e Tecnologia (FCT) and COMPETE 2020 program are gratefully acknowledged for funding this work with PTDC/EEI-AUT/5141/2014 (Automatic Adaptation of a Humanoid Robot Gait to Different Floor-Robot Friction Coefficients). c  Springer Nature Switzerland AG 2020 M. Botto-Tobar et al. (Eds.): ICAT 2019, CCIS 1194, pp. 53–64, 2020. https://doi.org/10.1007/978-3-030-42520-3_5