SPE & IMPERIAL COLLEGE SEMINAR: GETTING STARTED WITH DEEP LEARNING

Deep Learning (DL) is already bringing game-changing applications to the petroleum industry, and this is certainly the beginning of an enduring trend. Many petroleum engineers are interested to know more about DL but are not sure where to start. This one-day course aims to provide this introduction. The first half of the course presents the formalism of Logistic Regression, Neural Networks and Convolutional Neural Networks and some of their applications. In the afternoon, the environment associated with DL is discussed, from Python libraries to software repositories, important websites and big datasets available on the Web. The last part of the course is spent discussing the most promising subsurface applications of DL. This course, aimed at petroleum engineers and geoscientists, requires basic knowledge of Probability, Regression and Linear Algebra. 9.00-10.15: Supervised vs Unsupervised Learning, Regression, Logistic Regression. Basic theory and examples. 10.15-10.45: Break 10.45-12.15: Fully Connected Neural Networks and Convolutional Neural Networks. Basic theory and example 12.15-13.15: Lunch 13.15-14.30: Getting Practical: large existing datasets, websites, Python Deep learning libraries … 14.30-15.00: Break 15.00 -16.45: Four Deep Learning Industry Applications. 16.45-17.00: Conclusion The speakers Prof Olivier Dubrule is a Visiting Professor seconded by Total at the Department of Earth Science and Engineering at Imperial College London. He has spent most of his career developing new tools in relation with Geostatistics, Gaussian Processes and more recently Machine Learning. He is currently teaching the Machine Learning module of the MSc Advanced Computation Science and Engineering at Imperial College. Olivier has been the 2003 SEG/EAGE Distinguished Instructor, and has taught many courses through Professional Societies in the last twenty-five years. He has authored two books, one of which has been translated in Russian and Farsi. Lukas Mosser is a PhD Student in Machine Learning. Since he started his PhD in October 2016, Lukas has published important papers on Machine Learning in major peer-reviewed publications such as Transport in Porous Media and Physical Review Letters E. He specializes in the development of methods based on Generative Adversarial Networks (GANs) in relation to subsurface applications. Lukas has also obtained numerous awards in the Machine Learning hackathons in which he recently competed.   The Department in which we work: Today, the Department of Earth Science and Engineering (http://www.imperial.ac.uk/earth-science/) hosts approximately 50 academic staff, 70 research and support staff and 600 undergraduate and postgraduate students. It is ranked in first place as the top UK Earth Science department in the Guardian University League Tables 2019. Whether it is through undergraduate degrees in Geology, Geophysics, or Earth and Planetary Science, or our world leading Petroleum MSc courses, the Department is leading the way in Geoscience education and is committed to excellence in cutting-edge research.