Hungarian Machine Learning Days

2024. 07. 30 - 08. 01. 

About us

The Hungarian Machine Learning Days provide an opportunity for Hungarian machine learning researchers working in foreign institutions to meet each other and connect with those working in Hungary, including the younger generation and Ph.D. students.

The organization and a significant portion of the costs are covered by the National Artificial Intelligence Laboratory, established in 2020 (https://milab.hu/), whose main mission is to unite domestic researchers, organize events (such as www.ai-expo.hu), and build connections with European AI Centers and ELLIS (ellis.eu).

During the 3-day informal summer event, internationally recognized researchers will present tutorials, young researchers will give short presentations, posters, and ideas, and there will be many informal joint programs.

Participation, lunch, and coffee breaks are free of charge but require registration. On-site accommodation can be booked at the organizers at one's own expense (EUR 95.5 single 106 double per night). 

Mesterséges Intelligencia Nemzeti Labor (MILAB)

Programme - subject to change


All presentations are in English. Roundtables might be in Hungarian depending on participation. For posters, please apply during registration in time. Late poster applications risk no space remaining available. 


Magyar Machine Learning Találkozó

Day 2 - July 31

9:00-10:00 coffee, welcome

10:00 Csaba Szepesvári (Deepmind/Alberta): Tutorial: Lessons about scaling up reinforcement learning
11:00 Noémi Éltető (Deepmind): Planning with action sequence reuse
11:30 Gergő Flamich (Cambridge): An Introduction to Relative Entropy Coding and Its Applications
12:00 Botond Szabó (Bocconi University): Theoretical guarantees for variational Bayes
12:30-14:00 Lunch at the venue

Student talks


14:00 Anna Kerekes and Anna Mészáros and Szilvi Ujváry (ETH & Cambridge & Cambridge): Understanding LLMs Requires More Than Statistical Generalisation

Welcome address


14:45 Roland Jakab, CEO, Hungarian Research Network: The AI4Science initiative and the Research Hungary Grant

15:00 Panel: Women in ML
- Why did you become a researcher? Why did you choose ML/data science as topic?
- Why are women a minority in ML? Do you feel yourself underrepresented? Do you feel any discrimination, whether negative or positive?

16:15-18:00 coffee, poster section

Magyar Machine Learning Találkozó

Day 3 - August 1

9:00-10:00 coffee, welcome

Regular talks

10:00 Ferenc Huszár (Cambridge): Tutorial on Causal inference from Observational Data
11:00 Baráth Dániel (ETH Zürich): Applications of 3D Scene Graphs
11:30 Csaba Beleznai (AIT Austria): Robot perception from geometric cues
12:00 Attila Lovas (Rényi): On the Strong Stability of Ergodic Iterations

12:30-14:00 Lunch at the venue

Short/Student talks

14:00 Szabolcs Szentpéteri (HUN-REN SZTAKI): Sample Complexity of Distribution-Free Confidence Regions
14:15  Ambrus Tamás (HUN-REN SZTAKI): Data-Driven Upper Confidence Bounds for Symmetric Heavy-Tailed Bandits
14:30 Attila Csordás (AgeCurve, Cambridge, UK): Combining Blood Aging Insights: From Somatic Evolution to Single-Cell Atlases
14:45 Levente Halmosi, Bálint Mohos, Márk Jelasity: Evaluating the Adversarial Robustness of Semantic Segmentation: Trying Harder Pays Off

15:00
Panel: How to promote/help Hungarian students in ML

Posters

  • Ádám Fodor
    ELTE-IK
    Robust Emotion Sensing (work in progress)
  • Ádám Nagy
    Szegedi Tudományegyetem - Szoftverfejlesztési Tanszék
     
    The Actigraphy-Based Identification of Premorbid Latent Liability of Schizophrenia and Bipolar Disorder
  • Anh Tuan (László) Hoang
    HUN-REN SZTAKI
    IDENAS: Internal Dependency Exploration for Neural Architecture Search
  • Anna Gelencsér-Horváth
    ELTE IK
    Mobile Multi-Sensor Approaches for Assisting the Visually Impaired: Cybathlon tasks as use-case
  • Aron Fothi
    ELTE IK AI
    Cluster2Former
  • Balint Daroczy
    HUN-REN SZTAKI
    1. Length independent generalization bounds for deep SSM architectures with stability constraints (student poster by Daniel Racz, accepted at NGSM at ICML 2024)
    2. Geometrical aspects of continual learning and mode connectivity (student poster by Beatrix Benko)
  • Bálint Horváth
    HUN-REN SZTAKI, BME
    Nonparametric Simultaneous Confidence Bands with Paley-Wiener Kernels
  • Bence Gercuj
    HUN-REN SZTAKI
    Adaptive, Hybrid Feature Selection
  • Bernadett Vas
    HUN-REN Wigner Research Centre for Physics, Department of Computational Sciences
    Assessment of enhanced posterior expressivity via importance weighting in VAEs for natural images
  • Bruno Melicio
    Eötvös Loránd University
    Detecting hand-object interaction and child play in ADOS-2 videos
  • Domonkos Martos
    Wigner CSNL
    Scale Mixtures Variational Autoencoder: learning contrast invariant representation with an inductive bias 
  • Imre Molnár
    ELTE IK MIT NIPG
    FaceGym: Overcoming special needs related to facial expressions
  • Iván Fejes
    HUN-REN SZTAKI
    Age acceleration based on face photos predicts mortality and differs among occupations
    (Bence Király, Iván Fejes, Csaba Kerepesi)
  • János Adrián Gulyás
    Eötvös Loránd Tudományegyetem
    Keep Gesturing: A Game for Pragmatic Communication
  • Krisztián Buza
    Budapest Business University
    ROCKET with dynamic convolution for time series classification
  • Lehel Csato
    Babes-Bolyai University of Cluj-Napoca
    Wavelet-based prototype modelling
  • Márk Jelasity
    University of Szeged
    Levente Halmosi, Bálint Mohos, Márk Jelasity. Evaluating the Adversarial Robustness of Semantic Segmentation: Trying Harder Pays Off. Accepted for European Conference on Computer Vision (ECCV) 2024.
  • Miklós Máté Badó
    ELTE
    Keep Gesturing: A Game for Pragmatic Communication
  • Miklós Zorkóczy
    Legalaise Zorkóczy Law Office
    Legaltech tools in a paperless office
  • Simon Forrai
    HUN-REN SZTAKI
    Non-Linear Design of Experiments
  • Szabolcs Szentpéteri
    HUN-REN SZTAKI
    Sample Complexity of the Sign-Perturbed Sums Method
  • Szilárd Hunor Tóth
    HUN-REN SZTAKI
    Real Autonomous Vehicle Drifting Through Sim-to-Real Application of Reinforcement Learning Agents  
  • Szilvia Ujváry, Kerekes Anna, Mészáros Anna
    University of Cambridge
    Understanding LLMs Requires More Than Statistical Generalisation, ICML 2024 Position Paper
  • Tamás Takács
    ELTE IK, Department of Artificial Intelligence
    Scalable Distributed Reinforcement Learning in Multi-Agent Environments
  • Zsolt Csibi
    ELTE IK, Faculty of AI, NIPG
    The application of Large Language Models for in home physical rehabilitation
Registration

Registration: personalized invitation is sent to those on MILAB events list
 
Our location is the Budapesti Európai Ifjúsági Központ
(European Youth Centre Budapest)

Budapest, Zivatar Street 1, 1024

Budapesti Európai Ifjúsági Központ Budapest, Zivatar u. 1, 1024

Contact

Location

Budapesti Európai Ifjúsági Központ
(European Youth Centre Budapest)
Budapest, Zivatar u. 1, 1024
www.eycb.coe.int   

AI Website Maker