document

Contrastive Pessimistic Likelihood Estimation for Semi-Supervised Classification

Loog, M. (author)

Improvement guarantees for semi-supervised classifiers can currently only be given under restrictive conditions on the data. We propose a general way to perform semi-supervised parameter estimation for likelihood-based classifiers for which, on the full training set, the estimates are never worse than the supervised solution in terms of the log...

journal article 2016

document

Semi-supervised rail defect detection from imbalanced image data

Hajizadeh, S. (author), Nunez, Alfredo (author), Tax, D.M.J. (author)

Rail defect detection by video cameras has recently gained much attention in both<br/>academia and industry. Rail image data has two properties. It is highly imbalanced towards the non-defective class and it has a large number of unlabeled data samples available for semisupervised learning techniques. In this paper we investigate if positive...

conference paper 2016

document

Probabilistic defect-based risk assessment approach for rail failures in railway infrastructure

Jamshidi, A. (author), Faghih Roohi, S. (author), Nunez, Alfredo (author), Babuska, R. (author), De Schutter, B.H.K. (author), Dollevoet, R.P.B.J. (author), Li, Z. (author)

This paper develops a defect-based risk analysis methodology for estimating rail failure risk. The methodology relies on an evolution model addressing the severity level of rail surface defect, called squat. The risk of rail failure is assessed by analyzing squat failure probability using a probabilistic analysis of the squat cracks. For this...

conference paper 2016

document

Projected estimators for robust semi-supervised classification

Krijthe, J.H. (author), Loog, M. (author)

For semi-supervised techniques to be applied safely in practice we at least want methods to outperform their supervised counterparts. We study this question for classification using the well-known quadratic surrogate loss function. Unlike other approaches to semi-supervised learning, the procedure proposed in this work does not rely on...

journal article 2017

document

Model-based rare category detection for temporal data

Mandersloot, Jeroen (author)

Rare category detection is the task of discovering rare classes in unlabelled and imbalanced datasets. Existing algorithms focus almost exclusively on static data in which instances are assumed to be independent. In this thesis we propose an algorithm that is designed for temporal data. Specifically, we are interested in data with temporal...

master thesis 2018

document

Safe Semi-Supervised Learning

Bertazzi, Andrea (author)

Semi-supervised algorithms have been shown to possibly have a worse performance than the corresponding supervised model. This may be due to a violation of the assumptions on the data that are introduced in most classification systems. We study an approach that was previously shown to have guarantees of improvement for the LDA classifier in terms...

master thesis 2018

document

Towards deeper understanding of semi-supervised learning with variational autoencoders

Jurasiński, Karol (author)

Recently, deep generative models have been shown to achieve state-of-the-art performance on semi-supervised learning tasks. In particular, variational autoencoders have been adopted to use labeled data, which allowed the development of SSL models with the usage of deep neural networks. However, some of these models rely on ad-hoc loss additions...

master thesis 2019

document

Active Semi-Supervised Learning For Diffusions on Graphs

Das, Bishwadeep (author)

In statistical learning over large data-sets, labeling all points is expensive and time-consuming. Semi-supervised classification allows learning with very few labels. Naturally, selecting a few points to label becomes crucial as the performance relies heavily on the labeled points. The motivation behind active learning is to build an optimal...

master thesis 2019

document

Active Semi-Supervised Learning for Diffusions on Graphs

Das, Bishwadeep (author), Isufi, E. (author), Leus, G.J.T. (author)

Diffusion-based semi-supervised learning on graphs consists of diffusing labeled information of a few nodes to infer the labels on the remaining ones. The performance of these methods heavily relies on the initial labeled set, which is either generated randomly or using heuristics. The first sometimes leads to unsatisfactory results because...

conference paper 2020

document

A Distribution Dependent and Independent Complexity Analysis of Manifold Regularization

Mey, A. (author), Viering, T.J. (author), Loog, M. (author)

Manifold regularization is a commonly used technique in semi-supervised learning. It enforces the classification rule to be smooth with respect to the data-manifold. Here, we derive sample complexity bounds based on pseudo-dimension for models that add a convex data dependent regularization term to a supervised learning process, as is in...

conference paper 2020

document

Assumptions & Expectations in Semi-Supervised Machine Learning

Mey, A. (author)

The goal of this thesis is to investigate theoretical results in the field of semi-supervised learning, while also linking them to problems in related subjects as class probability estimation.<br/>

doctoral thesis 2020

document

Web-Based Economic Activity Classification: Comparing semi-supervised text classification methods to deal with noisy labels

Smalbil, Jos (author)

In order to provide accurate statistics for industries, the classification of enterprises by economic activity is an important task for national statistical institutes. The economic activity codes in the Dutch business register are less accurate for small enterprises since small enterprises are not labelled manually. To increase the quality of...

master thesis 2020

document

Partial discharges and noise classification under HVDC using unsupervised and semi-supervised learning

Morette, N. (author), Castro Heredia, L.C. (author), Ditchi, Thierry (author), Mor, A. R. (author), Oussar, Y. (author)

This paper tackles the problem of the classification of partial discharge (PD) and noise signals by applying unsupervised and semi-supervised learning methods. The first step in the proposed methodology is to prepare a set of classification features from the statistical moments of the distribution of the Wavelet detail coefficients extracted...

journal article 2020

document

A semi-supervised autoencoder framework for joint generation and classification of breathing

Pastor Serrano, O. (author), Lathouwers, D. (author), Perko, Z. (author)

Background and objective: One of the main problems with biomedical signals is the limited amount of patient-specific data and the significant amount of time needed to record the sufficient number of samples needed for diagnostic and treatment purposes. In this study, we present a framework to simultaneously generate and classify biomedical...

journal article 2021

document

Semi-Supervised Lane Detection With Deep Hough Transform

Lin, Y. (author), Pintea, S. (author), van Gemert, J.C. (author)

Current work on lane detection relies on large manually annotated datasets. We reduce the dependency on annotations by leveraging massive cheaply available unlabelled data. We propose a novel loss function exploiting geometric knowledge of lanes in Hough space, where a lane can be identified as a local maximum. By splitting lanes into separate...

conference paper 2021

document

Semi-Supervised Learning In Side-Channel Attacks

Slooff, Tom (author)

One of the most potent attacks against cryptographic implementations nowadays is side-channel attacks. Side-channel attacks use unintended leakages in the implementation, for example, electromagnetic radiation, to retrieve the secret key. Over time side-channel attacks have become more powerful, and recently the community has shifted towards...

master thesis 2021

document

Using cross-model learnings for the Gram Vaani ASR Challenge 2022

Patel, T.B. (author), Scharenborg, O.E. (author)

In the diverse and multilingual land of India, Hindi is spoken as a first language by a majority of its population. Efforts are made to obtain data in terms of audio, transcriptions, dictionary, etc. to develop speech-technology applications in Hindi. Similarly, the Gram-Vaani ASR Challenge 2022 provides spontaneous telephone speech, with...

journal article 2022

document

Deep Statistical Solver for Distribution System State Estimation

Habib, Benjamin (author)

Whereas in the past, Distribution Systems played a passive role in connecting customers to electricity, Distribution System Operators (DSOs) will have to take in the future a more active role in monitoring and regulating the network to deal with the new behaviors and dynamics of the system brought by the energy transition. State Estimation, a...

master thesis 2022

document

Evaluating the Effect of Data Drift on Model Performance

Wang, Wenhui (author)

Data drift refers to the variation in the production data compare to the training data and sometimes the machine learning model would decay because of it. Some machine learning models face the problem when in production: they receive drift data while there’s no ground truth to evaluate model performance, thus no way of determining the...

master thesis 2022

document

Developing health indicators for composite structures based on a two-stage semi-supervised machine learning model using acoustic emission data

Moradi, M. (author), Chiachío, Juan (author), Zarouchas, D. (author)

Composite structures are highly valued for their strength-to-weight ratio, durability, and versatility, making them ideal for a variety of applications, including aerospace, automotive, and infrastructure. However, potential damage scenarios like impact, fatigue, and corrosion can lead to premature failure and pose a threat to safety. This...

conference paper 2023