Mini-Symposium on introduction to Machine Learning
7th-8th September 2020 virtually and online (due to the COVID-19 pandemic).
The aim of this mini-symposium is to facilitate all ESRs high-level discussions on research challenges on machine learning in optical communications.
Day 1 – 7th September 2020
CET, Paris, Berlin, Rome time
09:20 – 9:30 Opening
09:30 – 11:30 Assoc. Prof. Darko Zibar, Dep. of Photonics Engineering, DTU FOTONIK. Introduction to machine learning and neural networks in particular. Apart from the lecture, the ESRs will need to spend some time implementing some simple neural networks.
11:30 – 11:45 Break
11:45 – 13:45 Assoc. Prof. Darko Zibar. Continuation of the talk.
13:45 – 15:00 Lunch
15:00 – 16:00 Dr. Kamalian-Kopae Morteza, Research Associate at the Institute of Photonics at Aston University. Machine learning-based equalisation in fibre-optic communication.
16:00 – 16:15 Break
16:15 – 17:15 Dr. Milad Sefidgaran, Postdoctoral fellow at Telecom Paris. Information theory of the optical fiber.
17:15 – 17:30 Closing
Day 2 – 8th September 2020
CET, Paris, Berlin, Rome time
14:20 – 14:30 Opening
14:30 – 15:30 Prof. David Saad, Professor in mathematics at Aston University. Machine learning beyond the hype – principled methods for photonics applications.
15:30 – 15:45 Break
15:45 – 16:45 Dr. Jelena Pesic & Dr. Matteo Lonardi, Nokia Bell Labs. Will Machine Learning mitigate the extra cost of increase in capacity?
16:45 – 17:00 Break
17:00 – 18:00 Prof. Nathan Kutz, Prof. Applied Mathematics, University of Washington. Machine Learning for Science: Data-Driven Discovery Methods for Governing equations, Coordinates and Sensors.
18:00 – 18:10 Closing
Assoc. Prof. Darko Zibar
Darko Zibar is Associate Professor at the Department of Photonics Engineering, Technical University of Denmark and the group leader of Machine Learning in Photonics Systems (M-LiPS) group. He received M.Sc. degree in telecommunication and the Ph.D. degree in optical communications from the Technical University of Denmark, in 2004 and 2007, respectively. He has been on several occasions (2006, 2008 and 2019) visiting researcher with the Optoelectronic Research Group led by Prof. John E. Bowers at the University of California, Santa Barbara, (UCSB). At UCSB, he has been working on topics ranging from analog and digital demodulation techniques for microwave photonics links and machine learning enabled ultra-sensitive laser phase noise measurements techniques. In 2009, he was a visiting researcher with Nokia-Siemens Networks, working on clock recovery techniques for 112 Gb/s polarization multiplexed optical communication systems. In 2018, he was visiting Professor with Optical Communication (Prof. Andrea Carena, OptCom) group, Dipartimento di Elettronica e Telecomunicazioni, Politecnico di Torino working on the topic of machine learning based Raman amplifier design. His resrearch efforts are currently focused on the application of machine learning technqiues to advance classical and quantum optical communication and measurement systems. Some of his major scientific contributions include: record capacity hybrid optical-wireless link (2011), record sensitive optical phase noise measurement technique that approaches the quantum limit (2019) and design of ultrawide band arbitrary gain Raman amplifier (2019). He is a recipient of Best Student paper award at Microwave Photonics Conference (2006), Villum Young Investigator Programme (2012), Young Researcher Award by University of Erlangen-Nurnberg (2016) and European Research Council (ERC) Consolidator Grant (2017). Finally, he was a part of the team that won the HORIZON 2020 prize for breaking the optical transmission barriers (2016).
Dr. Morteza Kamalian-Kopae
Morteza Kamalian-Kopae received his BSc in electrical engineering from Isfahan University of Technology, Isfahan, Iran, his MSc in communication engineering from Yazd University, Yazd, Iran, and his PhD in electrical engineering from Aston University, Birmingham, UK. Since graduation, he has been with Aston Institute of Photonic Technologies (AIPT) as a research fellow working on nonlinear Fourier transform, in particular, for periodic solutions of the nonlinear Schrödinger equation. His research interests include signal processing in optical communication, analysis of nonlinear dynamics, and wireless communication systems.
Dr. Milad Sefidgaran
Dr. Milad Sefidgaran obtained his BSc and MSc in Telecommunication from University of Tehran in 2007 and Sharif University of Technology in 2009, respectively. Then, he pursued his research on information theory and received his PhD from Telecom ParisTech in 2013. Between 2013 and 2015 he worked as postdoctoral associate at Telecom Paris and Sharif University of Technology. Later, he joined some Telecom companies as ZTE corporation and Huawei technologies. He is now working as postdoctoral researcher at Telecom Paris, since May 2019. His main research interests are information theory, optical fibers, wireless networks, and cellular technologies.
Prof. David Saad
Prof. David Saad holds the 50th Anniversary Chair of Complexity Physics at Aston University, Birmingham UK. He received a BA in Physics and a BSc in Electrical Engineering from the Technion, Haifa, Israel (1982), an MSc in Physics (1987) and a PhD in Electrical Engineering (1993) from Tel-Aviv University. He joined the Physics Department at the University of Edinburgh in 1992 and Aston University in 1995. His research, published in over 200 journal and conference papers, focuses on the application of methods from statistical physics and Bayesian statistics to a range of fields, which include neural networks, error-correcting codes, multi-node communication, network optimisation, routing, noisy computation, epidemic spreading and advanced inference methods.
Abstract presentation: Machine learning increasingly becomes a standard tool in a broad range of application domains. This mostly stems from their recent engineering successes, the availability of large quantities of data, and the easy access to optimised software packages and to more powerful machines. While many researchers are aware of deep multi-layer networks, due to their simplicity and the availability of dedicated software, machine learning is in fact an established field that also offers a variety of principled data-driven techniques for classification, regression, inference, optimisation and visual informatics. I will explain the broader context of machine learning methods, will review other leading machine learning techniques and motivate the use of principled probabilistic approaches for understanding and interpreting data. Additionally, I will refer to specific methods that are particularly useful for photonics applications.
Dr. Jelena Pesic
Dr. Jelena Pesic obtained her MSc degree at the University of Belgrade. Later she moved to France where in 2011 she obtained the Phd in Optical Network at the Université Bretagne Sud. Later, she worked as postdoctoral research at Telecom Paris, from 2011-2013, and at Inria, from 2013 to 2014. In 2014 Jelena joined Nokia Bell Labs as research engineering. Since Jun 2020 she is working as Systems Integration Specialist at Nokia Business Group, ION IP Optical Networks WDM technical expertise on the basis of the Nokia IP Transport product portfolio Research engineer in Nokia Bell labs OSA Ambassador (The Optical Society of America).
Dr. Matteo Lonardi
Dr. Matteo Lonardi obtained his BSc in Computer Engineering in 2012, his MSc in Communications Engineering in 2016 and his PhD in Information Technology (Optical Communications) at the Università degli Studi di Parma. Later he worked as researcher at the Unversità deli Studi di Parma, investigating the assessment, estimation, and monitoring of performance in dynamic modern optical communication systems developed to fight the expected capacity shortage. In March 2018 Matteo joined the Nokia Bell Labs as visiting researcher. Since November 2019 he is working as Research Engineer at Nokia Bell Labs.
Prof. Nathan Kutz
Professor Kutz was awarded the B.S. in Physics and Mathematics from the University of Washington in 1990 and the PhD in Applied Mathematics from Northwestern University in 1994. Following postdoctoral fellwoships at the Institute for Mathematics and its Applications (University of Minnesota, 1994-1995) and Princeton University (1995-1997), he joined the faculty of applied mathematics and served as Chair from 2007-2015.
Abstract presentation: Machine learning and artificial intelligence algorithms are now being used to automate the discovery of governing physical equations and coordinate systems from measurement data alone. However, positing a universal physical law from data is challenging: (i) An appropriate coordinate system must also be advocated and (ii) simultaneously proposing an accompanying discrepancy model to account for the inevitable mismatch between theory and measurements must be considered. Using a combination of deep learning and sparse regression, specifically the sparse identification of nonlinear dynamics (SINDy) algorithm, we show how a robust mathematical infrastructure can be formulated for simultaneously learning physics models and their coordinate systems. This can be done with limited data and sensors. We demonstrate the methods on a diverse number of examples, showing how data can maximally be exploited for scientific and engineering applications. The work also highlights the fact that the naive application of ML/AI will generally be insufficient to extract universal physical laws without further modification.