Neural networks have revolutionize many tasks in computer vision and interpretation, and will have significant rolein earth science tasks such as inversion, geological interpretation, prediction of earth parameters and water discovery.The talk will be divided into two parts. In the first part we will show that deep neural networks can be interpreted asdiscretizations of Partial Differential Equations that have been used for modeling for a very long time. This understandingcan help us design an appropriate network architecture that is task dependent and improve the network performance.In the second part of the talk, we discuss how such architectures can solve some problems in earth science, in particular, mapping of water and minerals, geological mapping, magnetic data segmentation and finding horizons in seismic data.
Speaker: Eldad Haber, Univ. of British Columbia
Contact:Website: Click to Visit
Save this Event:iCalendar
Windows Live Calendar
Stanford, CA 94305
Website: Click to Visit