Prediction of accelerator operation using machine learning
Yamanokuchi, T, Ando, S, Kinoshita, K, Bahadori, A & Kashiwao, T 2018, 'Prediction of accelerator operation using machine learning', IEEJ Transactions on Electrical and Electronic Engineering, vol. 13, pp. 656-657.
Published version available from:
This paper reports the regression analysis of the horizontal motion of a car driver’s foot from a brake pedal to an accelerator pedal using machine learning methods, namely, neural network (NN), radial basis function network (RBFN), random forest (RF), and support vector regression (SVR) models, in order to predict the timing of when the driver steps on the accelerator from an web camera movie. The approximation performances of these models are compared based on root mean squared error (RMSE) and the estimated timing when the driver’s foot reaches the accelerator pedal. From the results, the RF model could approximate the driver’s foot motion most precisely. © 2018 Institute of Electrical Engineers of Japan.