Conferences - Seminars

  10:30-11:30 GC C1 384

EDIC candidacy exam Exam president: Prof. Pascal Frossard Thesis advisor: Prof. Alexandre Massoud Alahi Co-examiner: Prof. Jean-Philippe Thiran Abstract Autonomous vehicles rely on an accurate perception module. One of the fundamental challenges is to efficiently track pedestrians surrounding a vehicle to anticipate risky situations. Over the past decades, researchers have formulated the tracking problem as ... Read more about "Multi-task learning for autonomous driving"
By George Adaimi
  14:00-16:00 BC 129

EDIC candidacy exam Exam president: Prof. Olivier Lévêque Thesis advisor: Prof. Michael Gastpar Co-examiner: Prof. Volkan Cevher Abstract There is an increasing concern that most current published research findings are false. The main cause seems to lie in the fundamental disconnection between theory and practise in data analysis. While the former typically relies on statistical ... Read more about "Adaptive Data Analysis and Error-Control via Information Measures"
By Amedeo Esposito
  12:00-13:00 RLC B1 520

Marc Robinson, professor at the Department of evolution and ecology at UNIL, will present the benefits from publishing preprints. Alain Foehn, PhD student at the Laboratory of hydraulic constructions at EPFL, will share his experience with submitting a paper to Elsevier. Read more about "NoonTalk#4: Publishing preprints"
  09:30-11:30 ELD 120

EDIC candidacy exam Exam president: Prof. Pascal Frossard Thesis advisor: Prof. Volkan Cevher Co-examiner: Prof. Alexandre Alahi Abstract Deep neural networks dominate the state-of-the-art models in tasks of computer vision, speech, bioinformatics etc. However, training modern deep neural networks is difficult, because of the highly non-convexity of the objective functions. On one hand, deep neural ... Read more about "Training and Tuning Deep Neural Networks: Faster, Stronger and Better"
By Chen Liu
  13:00-15:00 ELD 120

EDIC candidacy exam Exam president: Prof. Ali Sayed Thesis advisor: Prof. Volkan Cevher Co-examiner: Patrick Thiran Abstract In the recent years, there has been an increasing number of works that couples continuous-time processes with iterative descent methods in optimization literature to provide alternative perspective to convergence analysis and develop new algorithms. Background papers Coupling and ... Read more about "Differential Equations and Continuous-Time Dynamics in Optimization and Sampling"
By Ali Kavis
  09:00-11:00 INR 212

EDIC candidacy exam Exam president: Prof. Martin Rajman Thesis advisor: Prof. Pearl Pu Co-examiner: Prof. Wulfram Gerstner Abstract In the recent years, the successful application of sequence to sequence neural networks to statistical machine translation, including the usage of attention mechanism for the aim of translation alignment, has inspired a huge amount of work in ... Read more about "Modeling Emotional Dialogs with Sequence to Sequence Networks"
By Yubo Xie
  10:15 BC 420

Abstract Diagnosis and treatment methods based on big genomic data are becoming widespread as the cost of genome sequencing decreases gradually. In this context, disease-gene association studies are of great importance. The genomic data is very sensitive when compared to other data types and contains information about the individual and his / her relatives. Many ... Read more about "Achieving Scalable and Privacy-Preserving Whole Genome Variant Queries"
By Dr. Mete Akgün, University of Tübingen, Germany
  14:00-16:00 ELD 120

EDIC candidacy exam Exam president: Prof. Olivier Lévêque Thesis advisor: Prof. Volkan Cevher Co-examiner: Prof. Wulfram Gerstner Abstract We consider the problem of policy optimization in Reinforcement Learning via policy gradient method. We first show that it is possible to design a theoretical update rule for the policy parameters that converges to an optimal policy. ... Read more about "Policy gradient methods for Reinforcement Learning"
By Paul Rolland
  09:00-11:00 BC 329

EDIC candidacy exam Exam president: Prof. Michael Unser Thesis advisor: Prof. Martin Vetterli Co-examiner: Prof. Michael Gastpar Abstract Background papers Theory and Practice of Irregular Sampling (Part I only), by Feichtinger, H.G., Gröchening, K. Perfect Recovery and Sensitivity Analysis of Time Encoded Bandlimited Signals, by Lazar, A., Tóth, T. A Review of the Integrate-and-Fire Neuron ... Read more about "Sampling using Unsynchronized Time Encoding Machines"
By Karen Adam
  10:00-12:00 BC 329

EDIC candidacy exam Exam president: Prof. Mark Pauly Thesis advisor: Prof. Wenzel Jakob Co-examiner: Prof. Sabine Süsstrunk Abstract Background papers Read more about "Markov Chains for Light Transport"
By Merlin Nimier-David
  10:15 BC 420

Abstract In mix networks, communication anonymity is achieved by relaying messages through multiple routers called mixes, where each mix shuffles messages independently. Mix networks were originally set up as fixed sequences of mixes, called cascades. Later, more flexible configurations were proposed, such as free routes, where routes are not fixed and can be any sequence ... Read more about "Multiparty Routing: Secure Routing for Mixnets"
By Dr. Fatemeh Shirazi, KU Leuven, Belgium
  11:15-13:15 RLC D1 661

EDIC candidacy exam Exam president: Prof. Boi Faltings Thesis advisor: Prof. Volkan Cevher Thesis co-advisor: Prof. Pierre Dillenbourg Co-examiner: Dr. Mathieu Salzmann Abstract Background papers Apprenticeship Learning via Inverse Reinforcement Learning, by Pieter Abbel and Andrew Ng [ICML04] Model-free Imitation Learning with Policy Optimization, by Jonathan Ho, Jayesh Gupta and Stefano Ermon [ICML16] Understanding Black-box ... Read more about "Understanding the Behaviour of Inverse Reinforcement Learning Agents"
By Teresa Yeo
  14:00-16:00 BC 229

EDIC candidacy exam Exam president: Prof. Jean-Pierre Hubaux Thesis advisor: Prof. Bryan Ford Co-examiner: Prof. Katerina Argyraki Abstract The digitalization of most of today's world activities have given rise to a new class of decentralized and trustless systems. For one, distributed ledgers have found practical use in less than 10 years and have generated an ... Read more about "Improvements on Distributed Ledgers Applications and Self Sovereign Identity Systems"
By Nicolas Gailly
  09:30-11:30 BC 329

EDIC candidacy exam Exam president: Prof. Pascal Fua Thesis advisor: Prof. Wenzel Jakob Co-examiner: Prof. Mark Pauly Abstract Background papers Read more about "Learning Light Scattering Distributions"
By Delio Vicini
  11:00-13:00 ELD 120

EDIC candidacy exam Exam president: Prof. Emre Telatar Thesis advisor: Prof. Volkan Cevher Co-examiner: Prof. Daniel Kuhn Abstract Neural networks have shown great promise for task automation. However, the discovery of adversarial examples has upset this promise: small perturbations of the input data make the misclassification rate greatly increase. Recent approaches have been shown empirically ... Read more about "Obtaining Robust Classifiers via Regularization and Averaging Schemes."
By Fabian Latorre Gomez
  14:00-15:00 INR 113

We consider transmission over a cloud radio access network (CRAN) focusing on the framework of oblivious processing at the relay nodes (radio units), i.e., the relays are not cognizant of the users' codebooks. This approach is motivated by future wireless communications (5G and beyond) and the theoretical results connect to a variety of different information ... Read more about "Cloud Radio Access Networks, Distributed Information Bottleneck, and more: A Unified Information Theoretic View"
By Prof. Shlomo Shamai
The Viterbi Faculty of Electrical Engineering
Technion-Israel ...
  10:00-12:00 INN 326

EDIC candidacy exam Exam president: Prof. Edouard Bugnion Thesis advisor: Prof. George Candea Co-examiner: Prof. Christoph Koch Abstract Background papers Read more about "Formal Performance Contracts for Software Network Functions"
By Rishabh Iyer
  10:00-12:00 BC 329

EDIC candidacy exam (retake) Exam president: Prof. Wenzel Jakob Thesis advisor: Prof. Mark Pauly Co-examiner: Prof. Pierre Dillenbourg Abstract Computing a feasible disassembly sequence of parts for a given 3D structure is a fundamental research topic in geometric reasoning. Its inverse problem, which is to create the geometry of a 3D assembly according to predefined ... Read more about "A Graph-based Method for Understanding Interlocking Assemblies"
By Ziqi Wang
  14:15 BC 420

Abstract The introduction of computational resources at the network edge allows us to offload computation from clients and/or servers, thereby reducing response latency, backbone bandwidth, and computational requirements on clients. More fundamentally, edge-computing moves us from a Client-Server model to a Client-Edge-Server model. While this is an attractive paradigm for many applications, it raises the ... Read more about "CESSNA - Resilient Edge Computing"
By Dr. Yotam Harchol
University of California, Berkeley , USA
  09:30-11:30 BC 329

EDIC candidacy exam Exam president: Prof. Martin Jaggi Thesis advisor: Prof. Pascal Fua Thesis co-advisor: Dr. Mathieu Salzmann Co-examiner: Dr. François Fleuret Abstract Background papers Distilling the Knowledge in a Neural Network, by Hinton, G., Vinyals, O., Dean, J. Rethinking Feature Distribution for Loss Functions in Image Classification, by Wan, W., et al. Constrained Convolutional ... Read more about "Learning Compact Neural Networks"
By Shuxuan Guo
  10:00-12:00 BC 03

EDIC candidacy exam (retake) Exam president: Prof. Robert West Thesis advisor: Prof. Karl Aberer Co-examiner: Prof. Pierre Dillenbourg Abstract The cognitive dissonance phenomena states that people feel a mental discomfort when they are exposed to information that challenges their prior beliefs. This causes homophily which is socializing with people of same beliefs, and selective exposure, ... Read more about "Analyzing Political Polarization in Twitter"
By Elmas Tugrulcan
  10:15-11:00 SV 1717

Institute of Microengineering - Distinguished Lecture Abstract: Our insatiable demand for sensing, communication and computing has been a key driver in the relentless scaling of CMOS feature sizes. However, as the benefits of technology scaling are coming to an end, an increasing amount of out of the box thinking will be required to maintain the ... Read more about "Mixed-Signal Interfaces for the Data-Driven World"
  11:00-13:00 BC 229

EDIC candidacy exam Exam president: Prof. Boi Faltings Thesis advisor: Prof. Matthias Grossglauser Co-examiner: Prof. Martin Jaggi Abstract Background papers Read more about "Improving Neural Machine Translation through Interlingua"
By Aswin Suresh
  12:15-13:15 SV 1717

The vast majority of our knowledge on vibrio pathogenesis is based on the ancient and well-studied human pathogen, V. cholerae. Often neglected and relegated to specialized journals are vibrio infections in non-human species. While the studies of animal pathogens have benefited from the era of genomics, the search for pathogenesis determinants is often biased by ... Read more about "Environmental vibrios : « a walk on the wild side »"
By Frédérique Le Roux, Laboratory of Integrative Biology of Marine Models ...
  16:15-17:15 MXF 1

Crystal Structure Prediction (CSP) methods were developed on the assumption that an organic molecule would crystallize in its most stable crystal structure. Even implementing this approach is a challenge to computational chemistry methods,[1] as shown by the Cambridge Crystallographic Data Centre’s blind tests.[2] Polymorphism adds additional challenges, as this is usually a kinetic phenomenon with ... Read more about "Can we predict how pharmaceuticals will crystallize?"
By Prof. Sally Price (University College London, UK)
  10:15 BC 420

Today's global society strongly relies on collaborative document editing, which plays an increasingly large role in sensitive workflows. While other collaborative venues, such as secure messaging, have seen secure protocols being standardized and widely implemented, the same cannot be said for collaborative document editing. Popular tools such as Google Docs, Microsoft Office365 and Etherpad are ... Read more about "Capsule: A Protocol for Secure Collaborative Document Editing"
By Nadim Kobeissi
  09:00-12:30 CO 120

What strategies can I use to increase student participation? What potential advantages for student learning? At the end of this workshop, you will be able to: explain the importance of using interactive teaching strategies; select appropriate teaching strategies to achieve learning objectives; construct effective interactive teaching scenarios. Read more about "Effective Interactive Teaching"
By Siara Isaac
  10:15-11:30 BC 420

Abstract: Have you ever been stuck in meetings with discussions going in circles, meetings from which no decision emerges or one with which everyone is frustrated? Do you sometimes feel that you misunderstood the relationships among members of your professional or personal circles? Have you ever wondered how you could get to stop smoking, eat ... Read more about "Understanding decisions, relationships and behaviors through text… and Natural Language Processing"
By Léa Deleris
  10:00-12:00 INF 211

EDIC candidacy exam Exam president: Prof. Babak Falsafi Thesis advisor: Prof. Arjen Lenstra Thesis co-advisor: Prof. Serge Vaudenay Co-examiner: Prof. Bryan Ford Abstract Background papers Lattice Signatures and Bimodal Gaussians (BLISS), by Ducas, L., et al. Side-Channel Attacks on BLISS Lattice-Based Signatures -- Exploiting Branch Tracing Against strongSwan and Electromagnetic Emanations in Microcontrollers, by Espitau, ... Read more about "Side-channel evaluation of NIST's candidates for post-quantum standardisation"
By Aymeric Genet
  09:00-17:00 CO 120

This workshop puts participants to practice lecturing by presenting 5 minutes of their teaching. Advice and feedback is given based to develop scientific communication skills appropriate for teaching Read more about "Effective Lecturing"
By Ingrid Le Duc

Machine learning has come to play a prominent role in modern finance. Current applications range from text-based analysis of business reports to deep learning for credit risk and portfolio management. But the noise and behavioral elements inherent in financial data often require nonstandard machine learning solutions, possibly yet to be developed. The full potential of ... Read more about "Swissquote Conference 2018 on Machine Learning in Finance"
By Hans Buehler, JP Morgan - Rama Cont, University of Oxford - Isabelle ...
  09:00-17:00 EPFL, SG

The topic of the Swiss Inter- and Transdisciplinarity Day 2018 is "Inter- and Transdisciplinarity in a Digital World". With this day, we want to explore and reflect experiences and visions on how digitalisation can be devised for inter- and transdisciplinary research and education in a meaningful way. More information and call for contributions Read more about "Swiss Inter- and Transdisciplinarity Day 2018"