Synthetic data generation using WGAN-GP: Comparison between real and synthetic signals for all eight evaluated geometrical track irregularities

Masterarbeit fertiggestellt: Modelling and Synthesis of Track Irregularities for Data Augmentation using Advanced Schemes of Generative Adversarial Networks

1. Februar 2023

Autor: Guillermo Osio Arruti

In recent years, the world of Artificial Intelligence (AI) and more punctually Machine Learning (ML) has shed light on Generative Adversarial Networks (GANs). This algorithm is usually composed of two networks, a generator, and a discriminator. These will engage in an adversarial training where each will adjust its own parameters to perform better at a min-max optimization problem.

Weitere Informationen

Zum Seitenanfang