Geological Applications of Generative Adversarial Networks
Generative Adversarial Networks (GANs) are networks specialised to learn the distribution of a particular data set and hence are able to generate new patterns. They have recently found use in several different fields (e.g. They are used in AlphaFold developed in Google’s DeepMind to predict the 3D structure of Proteins). The project aim is to build on the previous progress of GANs applications to Geology and further study issues related to instabilities of training and shortage of data.
Dr Ahmed Elsheikh and Dr Uisdean Nicholson