J
Jakub M. Tomczak
Researcher at VU University Amsterdam
Publications - 129
Citations - 4012
Jakub M. Tomczak is an academic researcher from VU University Amsterdam. The author has contributed to research in topics: MNIST database & Restricted Boltzmann machine. The author has an hindex of 23, co-authored 116 publications receiving 2653 citations. Previous affiliations of Jakub M. Tomczak include University of Wrocław & University of Amsterdam.
Papers
More filters
Proceedings Article
Attention-based Deep Multiple Instance Learning
TL;DR: In this paper, a neural network-based permutation-invariant aggregation operator is proposed to learn the Bernoulli distribution of the bag label, where the bag-label probability is fully parameterized by neural networks.
Proceedings Article
VAE with a VampPrior
Jakub M. Tomczak,Max Welling +1 more
TL;DR: In this article, the VampPrior consists of a mixture distribution (e.g., a mixture of Gaussians) with components given by variational posteriors conditioned on learnable pseudo-inputs.
Journal ArticleDOI
Ensemble boosted trees with synthetic features generation in application to bankruptcy prediction
TL;DR: A novel ensemble model for bankruptcy prediction that utilizes Extreme Gradient Boosting for learning an ensemble of decision trees is proposed and a new approach for generating synthetic features to improve prediction is proposed.
Posted Content
Attention-based Deep Multiple Instance Learning
TL;DR: In this paper, a neural network-based permutation-invariant aggregation operator is proposed to learn the Bernoulli distribution of the bag label, where the bag-label probability is fully parameterized by neural networks.
Proceedings Article
Hyperspherical variational auto-encoders
TL;DR: In this article, a hyperspherical VAE (S-VAE) was proposed, which is more suitable for capturing data with a hyperspherespherical latent structure.