Projects
Semi-Supervised Models with Deep Architectures and Applications KU Leuven
In many real-life applications, ranging from data mining to machine perception, labeled data are very hard and expensive to attain as it requires expert human efforts. Therefore in many cases one often encounters a large number of unlabeled data whereas the labeled data are rare. Semi-Supervised Learning (SSL) is a framework in machine learning that aims at learning from both labeled and unlabeled data points and thus reducing the human ...
Learning signal representations of acoustic manifolds KU Leuven
Efficient analysis and processing of temporally and spatially sampled sound fields is paramount for many applications, including acoustic scene analysis, room geometry inference, and acoustic signal enhancement. Traditional approaches often rely on simplified physical and statistical models of the data and the data generation processes. With the advances in machine learning, prior models are being replaced by models learned from training ...
Flexible and weak supervision techniques for learning KU Leuven
Anomaly detection methods aim to identify examples that do not follow the expected behavior. For various reasons, anomaly detection is typically tackled by using unsupervised approaches that assign real-valued anomaly scores based on various heuristics. For instance, one can assume that anomalies fall in low-density regions and compute the negative log-likelihood as anomaly score.
Because anomaly scores are often hard to interpret, ...
Deep Learning in the Built Environment for Large-Scale Mapping and Detailed Land Use KU Leuven
Highly detailed, large-scale land-use monitoring so far in Belgium has been absent. Lots of data is available such as satellite and aerial imagery but also point cloud data from LiDAR. However, the handling of that data is either a manual interpretation task or a heavily supervised machine-learning method is used. The manual labeling of training data for machine-learning networks is a common approach, but highly intensive and costly. ...
Deep Restricted Kernel Machines: Methods and Foundations KU Leuven
This research proposal entitled "Deep Restricted Kernel Machines: Methods and Foundations" is related to two main directions in the field of machine learning:
- deep learning
- support vector machines and kernel methods
This project aims at an in-depth study of the recently proposed "Deep Restricted Kernel Machines" (Deep RKM). A method of conjugate feature duality is used to obtain a representation in terms ...
Supporting Pregnancy Care Based on Heterogeneous and Longitudinal Data Analysis KU Leuven
With the advent of information technology in healthcare, data occupy a central place in clinical practice. This digital dependency calls for efficient tools to assist physicians in managing and processing the explosion of data resources. These tools are commonly known as clinical decision support systems (CDSS). They cover a broad range of applications from alerting physicians of drug allergy to the facilitation of administrative data ...
Automated sleep analysis from wearable monitoring devices. KU Leuven
Sleep-wake disturbances are widespread across the disease spectrum. They are among the earliest and most disruptive symptoms in Alzheimer’s disease (AD), which is the number one neurodegenerative disease and represents a huge burden in our aging society. Sleep-wake disturbances are also associated with epilepsy, which in turn interacts with AD. While there is plenty of evidence for the interplay between AD, epilepsy and sleep, the nature of ...
Semi-automatically anomaly identification: the way forward to assurance over financial information Hasselt University
Deep, personalized epileptic seizure detection KU Leuven
There is an urgent need for objective epileptic seizure monitoring in the home environment. The only reliable way of monitoring seizures is by measuring brain activity, which is possible by means of the electroencephalogram (EEG). Miniaturized EEG devices with different electrode configurations that could be worn at home became available, but the challenge remains how to automatically detect the presence of seizures. Having demonstrated ...