Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox

Picture for Josif Grabocka

Transformers Can Do Bayesian Inference


Jan 25, 2022
Samuel Müller, Noah Hollmann, Sebastian Pineda Arango, Josif Grabocka, Frank Hutter

* Accepted at ICLR 2022 

  Access Paper or Ask Questions

Multi-task problems are not multi-objective


Oct 14, 2021
Michael Ruchte, Josif Grabocka


  Access Paper or Ask Questions

Regularization is all you Need: Simple Neural Nets can Excel on Tabular Data


Jun 21, 2021
Arlind Kadra, Marius Lindauer, Frank Hutter, Josif Grabocka


  Access Paper or Ask Questions

HPO-B: A Large-Scale Reproducible Benchmark for Black-Box HPO based on OpenML


Jun 11, 2021
Sebastian Pineda Arango, Hadi S. Jomaa, Martin Wistuba, Josif Grabocka


  Access Paper or Ask Questions

Efficient Multi-Objective Optimization for Deep Learning


Mar 24, 2021
Michael Ruchte, Josif Grabocka


  Access Paper or Ask Questions

Hyperparameter Optimization with Differentiable Metafeatures


Feb 07, 2021
Hadi S. Jomaa, Lars Schmidt-Thieme, Josif Grabocka


  Access Paper or Ask Questions

Few-Shot Bayesian Optimization with Deep Kernel Surrogates


Jan 19, 2021
Martin Wistuba, Josif Grabocka

* Published as a conference paper at ICLR 2021 

  Access Paper or Ask Questions

HIDRA: Head Initialization across Dynamic targets for Robust Architectures


Oct 28, 2019
Rafael Rego Drumond, Lukas Brinkmeyer, Josif Grabocka, Lars Schmidt-Thieme


  Access Paper or Ask Questions

Chameleon: Learning Model Initializations Across Tasks With Different Schemas


Oct 01, 2019
Lukas Brinkmeyer, Rafael Rego Drumond, Randolf Scholz, Josif Grabocka, Lars Schmidt-Thieme

* 9 pages, 6 figures 

  Access Paper or Ask Questions

Hyp-RL : Hyperparameter Optimization by Reinforcement Learning


Jun 27, 2019
Hadi S. Jomaa, Josif Grabocka, Lars Schmidt-Thieme


  Access Paper or Ask Questions

In Hindsight: A Smooth Reward for Steady Exploration


Jun 24, 2019
Hadi S. Jomaa, Josif Grabocka, Lars Schmidt-Thieme


  Access Paper or Ask Questions

Dataset2Vec: Learning Dataset Meta-Features


May 27, 2019
Hadi S. Jomaa, Josif Grabocka, Lars Schmidt-Thieme


  Access Paper or Ask Questions

Learning Surrogate Losses


May 24, 2019
Josif Grabocka, Randolf Scholz, Lars Schmidt-Thieme


  Access Paper or Ask Questions

Multi-Label Network Classification via Weighted Personalized Factorizations


Feb 25, 2019
Ahmed Rashed, Josif Grabocka, Lars Schmidt-Thieme

* Accepted in ICAART 2019 

  Access Paper or Ask Questions

Data-Driven Vehicle Trajectory Forecasting


Feb 09, 2019
Shayan Jawed, Eya Boumaiza, Josif Grabocka, Lars Schmidt-Thieme

* Published in ECML KNOWMe: 2nd International Workshop on Knowledge Discovery from Mobility and Transportation Systems 2018 

  Access Paper or Ask Questions

NeuralWarp: Time-Series Similarity with Warping Networks


Dec 20, 2018
Josif Grabocka, Lars Schmidt-Thieme


  Access Paper or Ask Questions

Channel masking for multivariate time series shapelets


Nov 02, 2017
Dripta S. Raychaudhuri, Josif Grabocka, Lars Schmidt-Thieme

* 12 pages 

  Access Paper or Ask Questions

Optimal Time-Series Motifs


May 03, 2015
Josif Grabocka, Nicolas Schilling, Lars Schmidt-Thieme

* Submitted to KDD2015 

  Access Paper or Ask Questions

Ultra-Fast Shapelets for Time Series Classification


Mar 17, 2015
Martin Wistuba, Josif Grabocka, Lars Schmidt-Thieme

* Preprint submitted to Journal of Data & Knowledge Engineering January 24, 2015 

  Access Paper or Ask Questions

Scalable Discovery of Time-Series Shapelets


Mar 11, 2015
Josif Grabocka, Martin Wistuba, Lars Schmidt-Thieme

* Under review in the journal "Knowledge and Information Systems" (KAIS) 

  Access Paper or Ask Questions

Time-Series Classification Through Histograms of Symbolic Polynomials


Dec 23, 2013
Josif Grabocka, Martin Wistuba, Lars Schmidt-Thieme


  Access Paper or Ask Questions

Invariant Factorization Of Time-Series


Dec 23, 2013
Josif Grabocka, Lars Schmidt-Thieme


  Access Paper or Ask Questions