Learn More
This paper applies the minimum gradient method (MGM) to denoise signals in engineering problems. The MGM is a novel technique based on the complexity control, which defines the learning as a bi-objective problem in such a way to find the best trade-off between the empirical risk and the machine complexity. A neural network trained with this method can be(More)
— This paper addresses the problem of inverting Ground Penetrating Radar (GPR) data, to find the buried inclusions characteristics depth and radii considering a non-homogenous host media by using neural networks (NN). The aim is the detection and characterization of inclusions in concrete structures. A novel asynchronous model is proposed to the NN(More)
— We are investigating the radar as a non-destructive technique to detect early-stage flaws and reinforcement in concrete structures. In this paper, we simulate the ground-penetrating radar (GPR) assessment of concrete structures using three-dimensional (3D) finite-difference time-domain (FDTD). The discussion focuses on the evaluation of the performance of(More)
  • 1