2021 - 2022

Past Talks

  • Date : 20 June, 2022
    Speaker :     Tomer Koren (Tel Aviv University)
    Title : "    Benign underfitting of Stochastic Gradient Descent, and some other peculiarities"

  • Date : 13 June, 2022
    Speaker :     Chloé Rouyer (University of Copenhagen)
    Title : "    A Near-Optimal Best-of-Both-Worlds Algorithm for Online Learning with Feedback Graphs"

  • Date : 30 May, 2022
    Speaker :     Andres Masegosa (Aalborg University)
    Title : "    Diversity and Generalization in Neural Network Ensembles"

  • Date : 9 May, 2022
    Speaker :     Stefan Sommer (University of Copenhagen)
    Title : "    Stochastic shape analysis and probabilistic geometric statistics"

  • Date : 2 May, 2022
    Speaker :     Clint Chin Pang Ho (City University of Hong Kong)
    Title : "    Robust Bellman Updates with s-Rectangularity"

  • Date : 21 March, 2022
    Speaker :     Julian Zimmert (Google Research)
    Title : "    Efficient Methods for Online Multiclass Logistic Regression"

  • Date : 14 March, 2022
    Topic : Internal seminar by Hippolyte Raymond Bourel

  • Date : 07 March, 2022
    Speaker :     Jalal Kazempour (DTU)
    Title : "    Regression Markets and Application to Energy Forecasting"

  • Date : 03 March, 2022
    Speaker :     Amir Yehudayoff (Technion)
    Title : "A characterization of multiclass learnability"

  • Date : 28 Feburary, 2022
    Speaker :     Krikamol Muandet (MPI-IS)
    Title : "Toward Reliable Machine Learning with Instruments"

  • Date : 09 Feburary, 2022
    Speaker :     Jalal Etesami (EPFL)
    Title : "Foundations of Causal Inference: Challenges and Opportunities"

  • Date : 10 January, 2022
    Speaker :     Yanlin Chen (Centrum Wiskunde & Informatica)
    Title : "    Quantum Algorithms and Lower Bounds for Linear Regression with Norm Constraints"

  • Date : 13 December, 2021 (Playing NeurIPS keynote)
    Speaker :     Gabor Lugosi (ICREA Research Professor at UPF and BSE Research Professor)
    Title : "    Do We Know How to Estimate the Mean?"

  • Date : 06 December, 2021
    Speaker :     Julian Zimmert (Google, Berlin)
    Title : "    The Pareto Frontier of model selection for general Contextual Bandits"

  • Date : 29 November, 2021
    Speaker :     Hippolyte Raymond Bourel (University of Copenhagen)
    Title : "    Model-Based Reinforcement Learning Exploiting State-Action Equivalence"

  • Date : 22 November, 2021
    Speaker :     Nicklas Werge (Sorbonne University)
    Title : "Invariant Policy Learning: A Causal Perspective"

  • Date : 15 November, 2021
    Speaker :     Sorawit Saengkyongam (University of Copenhagen)
    Title : "    Invariant Policy Learning: A Causal Perspective"

  • Date : 08 November, 2021
    Speaker :     Rasmus Pagh (University of Copenhagen)
    Title : "    Efficient differential privacy in the shuffle model"

  • Date : 25 October, 2021
    Speaker :     Niklas Pfister (University of Copenhagen)
    Title : "    Statistical Testing under Distributional Shifts: Applications in Offline Contextual Bandits"

  • Date : 11 October, 2021
    Speaker :     Yi-Shan Wu (University of Copenhagen)
    Title : "C    hebyshev-Cantelli PAC-Bayes-Bennett Inequality for the Weighted Majority Vote"

  • Date : 04 October, 2021
    T    opic : Discussion of selected papers from ICML and COLT (Part 3)

  • Date : 26 September, 2021
    T    opic : Discussion of selected papers from ICML and COLT (Part 2)

  • Date : 20 September, 2021
    T    opic : Discussion of selected papers from ICML and COLT (Part 1)

  • Date : 13 September, 2021
    Speaker :     Shinji Ito (NEC Corporation)
    Title : "    Parameter-Free Multi-Armed Bandit Algorithms with Hybrid Data-Dependent Regret Bounds"

  • Date : 06 September, 2021
    Speaker :     Ola Rønning (University of Copenhagen)
    Title :     "General and Integrated Stein Variational Inference with EinStein"

  • Date : 30 August, 2021
    Speaker :     Haipeng Luo (University of Southern California)
    Title : "    The best of both worlds: stochastic and adversarial episodic MDPs with unknown transition"