Combining SVD With wavelet transform in synthetic seismic signal denoising

Two Types of Neural Network Prediction Models and Time Complexity

“…RBF neural network to predict spent much less time than the linear adaptive network, linear adaptive neural network has great changes with the different error requirement, and the RBF network computing time is essentially the same or little change in the different error accuracy. When there are high requirements in error, the computing time of linear adaptive neural network prediction model is about 10.5ms; In the same calculation accuracy the maximum computation time of RBF neural network is 0.055ms, relative to the low-pass filter lagging 0.68s in article [6,7], the RBF neural network prediction calculation time can be ignored, satisfying the real-time control requirement.…”

“…, where  is learning rate. Widrow-Hoff learning rule could only train a single layer linear neural network; for multi-layer linear network, by using the superposition principle, we can design a single-layer linear neural network, owning the considerable performance [7] . Assuming that the center vector of RBF network hidden layer basis function and the normalizing parameters have been determined by offline.…”

“…After low-pass filter to filter the random signal, the law is manifested, as well as tilting passenger cars. For tilting Passenger Car actuation control system using the reference input signal, articles [6,7] did in-depth studies of forecasting models and prediction algorithms, applying the linear regression, linear adaptive neural network and RBF neural network respectively. The prediction is for that the real-time control of prediction model and algorithm time complexity is directly related to the real-time control.…”

“…The prediction is for that the real-time control of prediction model and algorithm time complexity is directly related to the real-time control. Articles [6,7] studied the two types of neural network models and algorithms, and this paper will research, analyze and calculate the complexity of spend time. Firstly, the paper gives these two types of neural network models and algorithms, and then is its time complexity analysis and calculation.…”