Giambattista Parascandolo

a.k.a. GB

I started my PhD in machine learning at the Max Planck Institute for Intelligent Systems and ETH Zürich in 2017, as a doctoral fellow of the Center for Learning Systems, where I am co-supervised by Bernhard Schölkopf and Thomas Hofmann. I am also an Ellis student.

During my PhD I was an intern at DeepMind and Google X.

I just want to understand and create intelligence. My main research interest is out-of-distribution generalization in deep learning.

News

  • 01/2021: Learning explanations that are hard to vary and A teacher-student framework to distill future trajectories have been accepted at ICLR 2021

  • 11/2020: A Seq2Seq approach to Symbolic Regression accepted at two NeurIPS 2020 Workshops: ”Learning Meets Combinatorial Algorithms” and ”Knowledge Representation & Reasoning Meets Machine Learning”

  • 08/2020: I (remotely) attended the MIT - Center for Brains Minds and Machines, Summer Course 2020

  • 02/2020: Started my year at ETH Zurich with Thomas Hofmann

  • 06/2019: Joined DeepMind in London for a summer internship, our project Divide-and-Conquer Monte Carlo Tree Search For Goal-Directed Planning is now on arXiv

  • 03/2019: Finally created this website, and added old news to make it look like it didn't take me 2 years to do it

  • 11/2018: Generalization in anti-causal learning accepted at the NeurIPS18 Workshop on Critiquing and Correcting Trends in Machine Learning

  • 10/2018: Adaptive Skip Intervals: Temporal Abstraction for Recurrent Dynamical Models accepted at NeurIPS 2018

  • 05/2018: Learning Independent Causal Mechanisms accepted at ICML 2018

  • 05/2018: Tempered Adversarial Networks accepted at ICML 2018

  • 04/2018: I will join Google X in Mountain View for a summer internship. By the end, I will have worked on a project exploring automated design for electromagnetic devices, with applications ranging from integrated photonics to metasurfaces, designed on ultra-large-scale simulators

  • 11/2017: Learning Independent Causal Mechanisms spotlight at NIPS 2017 workshop on Learning Disentangled Representations

  • 11/2017: ConvWave accepted at NIPS 2017 workshop on Deep Learning for Physical Sciences

  • 09/2017: Avoiding Discrimination Through Causal Reasoning accepted at NIPS 2017

  • 06/2017: Participated in the Machine Learning Summer School in Tübingen, as student and official master of games for social activities.

  • 06/2017: Participated in the Google Machine Learning Summer Summit at Google Zürich

  • 03/2017: Started my PhD at the Max Planck Insitute for Intelligent Systems in Tübingen and ETH Zürich

Selected Publications (for the full list Google Scholar)


Learning explanations that are hard to vary

Giambattista Parascandolo*, Alexander Neitz*, Antonio Orvieto, Luigi Gresele, Bernhard Schölkopf (* equal contribution)

ICLR 2021

[paper] [code]


A teacher-student framework to distill future trajectories

Alexander Neitz*, Giambattista Parascandolo*, Bernhard Schölkopf (* equal contribution)

ICLR 2021

[paper]


A Seq2Seq approach to Symbolic Regression

Luca Biggio, Tommaso Bendinelli, Aurelien Lucchi, Giambattista Parascandolo

NeurIPS 2020 Workshop ”Learning Meets Combinatorial Algorithms”
NeurIPS 2020 Workshop ”Knowledge Representation & Reasoning Meets Machine Learning”

[paper]


Divide-and-Conquer Monte Carlo Tree Search For Goal-Directed Planning

Giambattista Parascandolo*, Lars Buesing*, Josh Merel, Leonard Hasenclever, John Aslanides, Jessica B. Hamrick, Nicolas Heess, Alexander Neitz, Theophane Weber (* equal contribution)

submitted

[paper] [supporting-website]


Adaptive Skip Intervals: Temporal Abstraction for Recurrent Dynamical Models

Alexander Neitz, Giambattista Parascandolo, Stefan Bauer, Bernhard Schölkopf

NeurIPS 2018

[paper]

Generalization in anti-causal learning

Niki Kilbertus*, Giambattista Parascandolo*, Bernhard Schölkopf (* alphabetical order)

NeurIPS 2018 Workshop on Critiquing and correcting trends in machine learning

[paper]

Learning Independent Causal Mechanisms

Giambattista Parascandolo, Niki Kilbertus, Mateo Rojas-Carulla, Bernhard Schölkopf

ICML 2018

also at: NIPS 2017 Workshop on Learning Disentangled Representations - Spotlight presentation

[paper] [bibtex]

Tempered Adversarial Networks

M. S. M. Sajjadi, Giambattista Parascandolo, Arash Mehrjou, Bernhard Schölkopf

ICML 2018

also at: ICLR 2018 Workshop

[paper]

Avoiding Discrimination Through Causal Reasoning

Niki Kilbertus, Mateo Rojas-Carulla, Giambattista Parascandolo, Moritz Hardt, Dominik Janzing, Bernhard Schölkopf

NeurIPS 2017

[paper] [bibtex] [poster] [in the press: MPI]

ConvWave: Searching for Gravitational Waves with Fully Convolutional Neural Nets

Timothy Gebhard*, Niki Kilbertus*, Giambattista Parascandolo, Ian Harry, Bernhard Schölkopf (* equal contribution)

NeurIPS 2017 Workshop on Deep Learning for Physical Sciences

[paper] [bibtex] [code] [poster]

Sound Event Detection in Multichannel Audio Using Spatial and Harmonic Features,

Sharath Adavanne, Giambattista Parascandolo, Pasi Pertilä, Toni Heittola, Tuomas Virtanen

Workshop on Detection and Classification of Acoustic Scenes and Events (DCASE) 2016 - 1st place in the challenge.

[paper]

Acoustic Scene Classification Using Convolutional Neural Networks

Michele Valenti, Aleksandr Diment, Giambattista Parascandolo, Stefano Squartini, Tuomas Virtanen

International Joint Conference on Neural Networks (IJCNN) 2017 - Best Student Poster Intel Award

IEEE AASP Challenge on Detection and Classification of Acoustic Scenes and Events (DCASE2016)

[paper]

Low-latency sound source separation using deep neural networks

Gaurav Naithani, Giambattista Parascandolo, Tom Barker, Niels Henrik Pontoppidan, Tuomas Virtanen

IEEE Global Conference on Signal and Information Processing (GlobalSIP) 2016

[paper]

Convolutional Recurrent Neural Networks for Polyphonic Sound Event Detection

Emre Cakir*, Giambattista Parascandolo*, Toni Heittola, Heikki Huttunen, Tuomas Virtanen (*equal contribution)

IEEE/ACM Transactions on Audio, Speech, and Language Processing (Journal)

[paper]

Recurrent neural networks for polyphonic sound event detection in real life recordings

Giambattista Parascandolo, Heikki Huttunen, Tuomas Virtanen

ICASSP 2016 - Best student paper of its special session

[paper]

Miscellaneous

  • Co-organized the Hackathon at the CLS Max Planck - ETH retreat 2018, and photographer for the 2017 retreat in the Alps

  • Reviewed for NeurIPS, ICML, ICLR

  • Coding and technical interviews for PhD candidates applying to the CLS and IMPRS programs

  • In my free time I play the piano, volleyball, football, volleypong, and with words. I also like to collect photons on film and DSLR sensors.

Contact

[first] [dot] [last] [at] tue.mpg [dot] de