University of Basel

Department of
Mathematics and Computer Science

Novel Trends in Data Science 2023

Monday, October 23, 2023

TimeSpeakerTitle
08.30 - 08.45Irene SchmidliWelcome Address Congressi Stefano Franscini
08.45 - 09.00Heike Freiberger, Heiko SchuldtIntroduction and Welcome
09.00 - 09.45Keynote: Sonali Parbhoo (Imperial College, London, UK)Mind your (Evaluation) Metrics: Towards Actionable Off-Policy Evaluation for Safety Critical Settings
09.45 - 10.10Amirehsan Khorashadizadeh, Valentin Debarnot and Ivan DokmanićDeep Local Transformers for Computed Tomography
10.10 - 10.35Jim Zhao, Aurelien Lucchi and Nikita DoikovStochastic Coordinate Cubic Regularized Newton for non-convex optimization
10.35 - 11.00Coffee Break
11.00 - 11.25Marcus J. Grote, Omar Lakkis and Carina S. SantosA Posteriori Error Estimates For The Wave Equation
11.25 - 11.50Navish Kumar, Aurelien Lucchi, Thomas Mollenhöff and Emtiyaz KhanNatural gradient square root variational inference
11.50 - 12.15Enrico Giudice, Jack Kuipers and Giusi MoffaA Bayesian Take on Gaussian Process Networks
12.15 - 14.00Lunch Break
15.00 - 17.00Hiking Tour
19.00Dinner

Tuesday, October 24, 2023

TimeSpeakerTitle
09.00 - 09.45Keynote: Julia Vogt (ETH Zürich)Multimodal Machine Learning in Medicine
09.45 - 10.10Federico Ravenda, Seyed Ali Bahrainian and Fabio CrestaniA Lightweight Self-Supervised Approach to do Topic Modelling
10.10 - 10.35Fynn Firouz Faber and Raphael WaltenspülUnderstanding Data Modeling Needs: Information Retrieval in the GLAM Sector
10.35 - 11.00Coffee Break
11.00 - 11.25Valentyna Pavliv and Isabel WagnerDark Patterns in Smart Toys: automated discovery and remediation
11.25 - 11.50Thomas Jakobsche and Florina M. CiorbaBridging the Gap: From Queuing to Execution in High Performance Computing Systems
11.50 - 12.15Martin Vahlensieck, Marco Vogt and Heiko SchuldtTowards Adaptive Data Management Using Containerization
12.15Lunch
14.00 - 14.45Keynote: Mahnaz Parian-Scherb (Roche)Pixels to Patients: Exploring Computer Vision's Impact on Healthcare
14.45 - 15.10Sanaullah Sanaullah, Shamini Koravuna, Ulrich Rückert and Thorsten JungeblutA Novel Spike Vision Approach for Robust Multi-Object Detection using SNNs
15.10 - 15.35Marcello Massimo Negri, Fabricio Arend Torres and Volker RothObjective Bayesian Lasso with Conditional Flows
15.35 - 16.00Emanuele Francazi, Aurelien Lucchi and Marco Baity JesiInitial Guessing Bias: How Untrained Networks Favor Some Classes
16.00 - 16.30Coffee Break
16.30 - 16.55Shiva Parsarad and Isabel WagnerPrivacy in Brain Computer Interfaces
16.55 - 17.20David Lengweiler, Marco Vogt and Heiko SchuldtTowards a Unified Data Integration, Access and Analysis Platform
17.20 - 17.45Alexander AhnEvaluating the Performance of Machine Learning Methods for Predicting Mortality in Intensive Care Unit Patients
17.45 - 18.10Rustem Islamov, Xun Qian, Slavomír Hanzely, Mher Safaryan and Peter RichtárikDistributed Newton-Type Methods with Communication Compression and Bernoulli Aggregation
18.10 - 18.35Dilana Irvine Rauch, Florian Klaus Kaiser and Frank SchultmannLocal Price Effects of Flood Exposure and Flood Risk in the Residential Property Market
19.00Social Dinner

 

Wednesday, October 25, 2023

TimeSpeakerTitle
09.00 - 09.45Keynote: Ilir Fetai, SBB (Switzerland)ML Model != ML Application
09.45 - 10.10Nima Akbari and Isabel WagnerPrivacy and Transparency for Virtual Reality
10.10 - 10.35Rahel Arnold and Heiko SchuldtExploring Mixed Reality Query Generation for Enhanced Multimedia Retrieval
10.35 - 11.00Coffee Break
11.00 - 11.25Yiming Wu, Heiko Schuldt and Marco VogtIntegrating DuckDB into Polypheny-DB
11.25 - 11.50Tim De Ryck, Ameya D. Jagtap and Siddhartha MishraOn the stability of extended and conservative physics-informed neural networks (XPINNs and cPINNs)
11.50 - 12.15Jonas Linkerhägner, Cheng Shi and Ivan DokmanićInterpreting Graph Denoising via Resonance
12.15 - 12.40Monika Nagy-Huber and Volker RothPhysics-Informed Boundary Integral Networks (PIBI-Nets): Data Assimilation for the Heat Equation
12.40 - 13.00NTDS OrganizersNTDS Debriefing

 

To top