Deep Learning for Time Series Classification – Applications to Surgical Data Science
Invited speaker: Germain Forestier
Summary: In recent years, deep learning approaches have demonstrated a tremendous success in multiple domains like image processing, computer vision or speech recognition. In this talk, I will review recent advances in deep learning for univariate and multivariate time series classification. I will present experimental results obtained with the principal architectures proposed in the literature. I will also discuss the main challenges linked with the use of deep learning like transfer learning and data augmentation. Finally, I will present some applications in the field of Surgical Data Science which is an emerging field with the objective of improving the quality of interventional healthcare and its value through capturing, organizing, analyzing, and modeling of data.
Short bio: Dr Germain Forestier received his PhD in Computer Science from the University of Strasbourg (France) in 2010. He then spent one year as a post-doctoral fellow at INRIA Rennes (France) / INSERM (French National Institute for Medical and Health Research), where he worked on biomedical data analysis. Since September 2011, he hold a position of Associate Professor at the University of Haute-Alsace (France) and he is a member of the IRIMAS research laboratory. He is also Senior Lecturer (Adjunct) at the Monash University (Australia). His research interests include data science, data mining, time series, machine learning, big data, artificial intelligence and deep learning.