no code implementations • 3 Apr 2024 • Boje Deforce, Meng-Chieh Lee, Bart Baesens, Estefanía Serral Asensio, Jaemin Yoo, Leman Akoglu
A two-fold challenge for TSAD is a versatile and unsupervised model that can detect various different types of time series anomalies (spikes, discontinuities, trend shifts, etc.)
1 code implementation • 16 Oct 2023 • Manon Reusens, Philipp Borchert, Margot Mieskes, Jochen De Weerdt, Bart Baesens
This paper investigates the transferability of debiasing techniques across different languages within multilingual models.
1 code implementation • 10 Oct 2023 • Jonathan Tonglet, Manon Reusens, Philipp Borchert, Bart Baesens
The performance of In-Context Learning depends heavily on the selection procedure of the supporting exemplars, particularly in the case of HybridQA, where considering the diversity of reasoning chains and the large size of the hybrid contexts becomes crucial.
1 code implementation • 16 Jul 2023 • Elena Tiukhova, Emiliano Penaloza, María Óskarsdóttir, Bart Baesens, Monique Snoeck, Cristián Bravo
We compare the results of various models to demonstrate the importance of capturing graph representation, temporal dependencies, and using a profit-driven methodology for evaluation.
no code implementations • 9 May 2023 • Boje Deforce, Bart Baesens, Jan Diels, Estefanía Serral Asensio
IoT data is a central element in the successful digital transformation of agriculture.
1 code implementation • 15 Nov 2022 • Elena Tiukhova, Emiliano Penaloza, María Óskarsdóttir, Hernan Garcia, Alejandro Correa Bahnsen, Bart Baesens, Monique Snoeck, Cristián Bravo
Leveraging network information for prediction tasks has become a common practice in many domains.
no code implementations • 3 Jun 2022 • Toon Vanderschueren, Robert Boute, Tim Verdonck, Bart Baesens, Wouter Verbeke
This work proposes to relax both assumptions by learning the effect of maintenance conditional on a machine's characteristics from observational data on similar machines using existing methodologies for causal inference.
no code implementations • 9 Feb 2022 • Toon Vanderschueren, Bart Baesens, Tim Verdonck, Wouter Verbeke
A central problem in business concerns the optimal allocation of limited resources to a set of available tasks, where the payoff of these tasks is inherently uncertain.
no code implementations • 13 Oct 2021 • Pieter De Koninck, Klaas Nelissen, Seppe vanden Broucke, Bart Baesens, Monique Snoeck, Jochen De Weerdt
Within the field of process mining, several different trace clustering approaches exist for partitioning traces or process instances into similar groups.
1 code implementation • 15 Sep 2020 • María Óskarsdóttir, Waqas Ahmed, Katrien Antonio, Bart Baesens, Rémi Dendievel, Tom Donas, Tom Reynkens
Finally, we combine these network features with the claim-specific features and build a supervised model with fraud in motor insurance as the target variable.
1 code implementation • 5 May 2020 • Sebastiaan Höppner, Bart Baesens, Wouter Verbeke, Tim Verdonck
Fraud detection is to be acknowledged as an instance-dependent cost-sensitive classification problem, where the costs due to misclassification vary between instances, and requiring adapted approaches for learning a classification model.
Applications
no code implementations • 3 May 2020 • Sam Verboven, Jeroen Berrevoets, Chris Wuytens, Bart Baesens, Wouter Verbeke
However, few data-driven tools that support strategic decision-making are available.
1 code implementation • 22 Mar 2020 • Bart Baesens, Sebastiaan Höppner, Irene Ortner, Tim Verdonck
Detecting fraud in such a highly imbalanced data set typically leads to predictions that favor the majority group, causing fraud to remain undetected.
no code implementations • 23 Feb 2020 • María Óskarsdóttir, Cristián Bravo, Carlos Sarraute, Jan Vanthienen, Bart Baesens
In terms of profit, the best model is the one built with only calling behavior features.
no code implementations • 3 Feb 2020 • Tine Van Calster, Filip Van den Bossche, Bart Baesens, Wilfried Lemahieu
This paper aims to facilitate this process for high-level tactical sales forecasts by comparing a large array of techniques for 35 times series that consist of both industry data from the Coca-Cola Company and publicly available datasets.
no code implementations • 21 Dec 2017 • Sebastiaan Höppner, Eugen Stripling, Bart Baesens, Seppe vanden Broucke, Tim Verdonck
Customer retention campaigns increasingly rely on predictive models to detect potential churners in a vast customer base.