Projects
New Methods for Self-Supervised Image Representation Learning KU Leuven
Representation learning plays a key role in most machine learning algorithms. When solving a particular task, the input is usually mapped to an intermediary latent space before being mapped to the final output space. We refer to the representation of an input as the set of feature values to which the input is mapped in the latent space. A good representation captures the essential information present in the data in a readily accessible ...
Modular representation theory of the periplectic Brauer algebra. Ghent University
Representation theory of the symmetric group is related to representation theory of the general linear group via Schur-Weyl duality. Similarly, Schur-Weyl duality also relates the orthogonal Lie group, the symplectic Lie group and the encompassing orthosymplectic Lie supergroup to the Brauer algebra, and it relates the periplectic Lie supergroup to the periplectic Brauer algebra. The goal of the project is to develop modular representation ...
Unsupervised Representation Learning and Health Insurance Anomaly Detection KU Leuven
One of the most important reasons behind the advances in machine learning is due to their increasingly powerful representation learning capabilities. The most well-known type of representation learning is a neural network, which is a powerful type of model that can automatically extract features out of input data. However, one has little control over the types of features that a neural network learns. This motivates research into learning ...
Medial temporal representation of word meaning embedded in natural language processing KU Leuven
Sentence comprehension deficits can arise despite preserved single word comprehension due to a deficit of integrating information into an overall meaning representation. In this proposal, I will leverage the rapid advances in Natural Language Processing (NLP) in computational sciences to investigate the neurobiology of natural language comprehension. Five fMRI and two eye tracking experiments will be conducted to examine how the brain ...
Improving language understanding through common sense and anticipation enriched representation learning KU Leuven
Trying to see whether various downstream language understanding tasks can be improved upon using representations of the symbolic words that are 'anticipatory'.
The intended properties of these 'anticipatory representations' are:
- High-level: they should represent some high level semantic concept, e.g. a spatial/temporal relation between words
- Common sense: they should bring 'common sense', which humans get from living in ...
The literary representation of the economic in the German novel, 1855 # 1901 KU Leuven
This dissertation presents a new perspective for the study of the German bourgeois realistic prose, by parallelizing literary poetics, financial epistemology, commercial procedures of observation and registration. Based on a corpus of literary programmes, economic studies, commercial and financial manuals, a poetology of credit-economic knowledge is first established. The nexus of these apparently very different paradigms of aesthetics and ...
Beyond Baudelaire's indignation : an enquiry into the interplay between the art markets and the decline of the representation theory in art criticism. The case of Brussels, 1848-1914. University of Antwerp
Translating the Real to the Reel: The Representation of Latinx Migration in Documentaries KU Leuven
The present doctoral study aims to chart the representation of Latinx migration to the US in documentary films. As US immigration policy grew increasingly restrictive over the course of the 20th century, public discourse became more antagonizing towards Latinx immigrants. Hence, there is a wealth of literature on stereotyping and antagonizing media frames of Latinxs. There is little information, however, on attempts by the media to approach ...
Representation Learning for Automated Document Classification KU Leuven
Classification of text documents into a hierarchical taxonomy of categories is
a challenging task in a multitude of subject domains. With an increase in
digitalization, the need for high performing automatic classification systems is
growing every day. In this dissertation, contributions are presented that take
up these challenges in the fields of online news articles and electronic healthcare
documents, ...