Matthijs van Leeuwen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Matthijs van Leeuwen has a total of 109 co-authors. Here are the top co-authors by number of shared publications:
- Zhong Li
Leiden Institute of Advanced Computer Science (LIACS), Leiden University, Leiden, The Netherlands
11 shared publications - Luc De Raedt
KU Leuven
10 shared publications - Lincen Yang
Leiden University
7 shared publications Arno Siebes
Universiteit Utrecht, Utrecht, Netherlands
6 shared publications- Mart P. Janssen
Transfusion Technology Assessment, Department of Research Sanquin Blood Supply Foundation Amsterdam The Netherlands
6 shared publications - Vladimir Dzyuba
KU Leuven
6 shared publications
Latest Publications
- 17 September 2026
- Published by Springer Nature
- p. 259-275
Characterizing Fundamental Differences Between Tabular Synthetic Data Generation Methods
- 17 September 2026
- Published by Springer Nature
- p. 259-275
Characterizing Fundamental Differences Between Tabular Synthetic Data Generation Methods
- Open Access
- 3 July 2026
- Published by Springer Nature in Data Mining and Knowledge Discovery
- Vol. 40 (5)
Graph neural networks based log anomaly detection and explanation
Event logs are widely used to record the status of high-tech systems, making log anomaly detection important for monitoring those systems. Most existing log anomaly detection methods take a log event count matrix or log event sequences as input, exploiting quantitative and/or sequential relationships between log events to detect anomalies. However, only considering quantitative or sequential relationships may result in low detection accuracy. To alleviate this problem, we propose a graph-based method for unsupervised log anomaly detection, dubbed Logs2Graphs , which first converts event logs into attributed, directed, and weighted graphs, and then leverages graph neural networks to perform graph-level anomaly detection. Specifically, we introduce One-Class Digraph Inception Convolutional Networks, abbreviated as OCDiGCN, a novel graph neural network model for detecting graph-level anomalies in a collection of attributed, directed, and weighted graphs. By integrating graph representation and anomaly detection, OCDiGCN learns a specialized representation that leads to high detection accuracy. Crucially, we furnish a concise set of nodes pivotal in OCDiGCN’s prediction as explanations for each detected anomaly, offering valuable insights for subsequent root cause analysis. Experiments on five benchmark datasets show that Logs2Graphs exhibits comparable or superior performance when compared to state-of-the-art log anomaly detection methods.
- 3 July 2026
- Published by Springer Nature in Data Mining and Knowledge Discovery
- Vol. 40 (5)
Open AccessGraph neural networks based log anomaly detection and explanation
Event logs are widely used to record the status of high-tech systems, making log anomaly detection important for monitoring those systems. Most existing log anomaly detection methods take a log event count matrix or log event sequences as input, exploiting quantitative and/or sequential relationships between log events to detect anomalies. However, only considering quantitative or sequential relationships may result in low detection accuracy. To alleviate this problem, we propose a graph-based method for unsupervised log anomaly detection, dubbed Logs2Graphs , which first converts event logs into attributed, directed, and weighted graphs, and then leverages graph neural networks to perform graph-level anomaly detection. Specifically, we introduce One-Class Digraph Inception Convolutional Networks, abbreviated as OCDiGCN, a novel graph neural network model for detecting graph-level anomalies in a collection of attributed, directed, and weighted graphs. By integrating graph representation and anomaly detection, OCDiGCN learns a specialized representation that leads to high detection accuracy. Crucially, we furnish a concise set of nodes pivotal in OCDiGCN’s prediction as explanations for each detected anomaly, offering valuable insights for subsequent root cause analysis. Experiments on five benchmark datasets show that Logs2Graphs exhibits comparable or superior performance when compared to state-of-the-art log anomaly detection methods.
- Open Access
Learning Subgroups with Maximum Treatment Effects without Causal Heuristics
Discovering subgroups with the maximum average treatment effect is crucial for targeted decision making in domains such as precision medicine, public policy, and education. While most prior work is formulated in the potential outcome framework, the corresponding structural causal model (SCM) for this task has been largely overlooked. In practice, two approaches dominate. The first estimates pointwise conditional treatment effects and then fits a tree on those estimates, effectively turning subgroup estimation into the harder problem of accurate pointwise estimation. The second constructs decision trees or rule sets with ad-hoc 'causal' heuristics, typically without rigorous justification for why a given heuristic may be used or whether such heuristics are necessary at all. We address these issues by studying the problem directly under the SCM framework. Under the assumption of a partition-based model, we show that optimal subgroup discovery reduces to recovering the data-generating models and hence a standard supervised learning problem (regression or classification). This allows us to adopt any partition-based methods to learn the subgroup from data. We instantiate the approach with CART, arguably one of the most widely used tree-based methods, to learn the subgroup with maximum treatment effect. Finally, on a large collection of synthetic and semi-synthetic datasets, we compare our method against a wide range of baselines and find that our approach, which avoids such causal heuristics, more accurately identifies subgroups with maximum treatment effect. Our source code is available at https://github.com/ylincen/causal-subgroup.
Open AccessLearning Subgroups with Maximum Treatment Effects without Causal Heuristics
Discovering subgroups with the maximum average treatment effect is crucial for targeted decision making in domains such as precision medicine, public policy, and education. While most prior work is formulated in the potential outcome framework, the corresponding structural causal model (SCM) for this task has been largely overlooked. In practice, two approaches dominate. The first estimates pointwise conditional treatment effects and then fits a tree on those estimates, effectively turning subgroup estimation into the harder problem of accurate pointwise estimation. The second constructs decision trees or rule sets with ad-hoc 'causal' heuristics, typically without rigorous justification for why a given heuristic may be used or whether such heuristics are necessary at all. We address these issues by studying the problem directly under the SCM framework. Under the assumption of a partition-based model, we show that optimal subgroup discovery reduces to recovering the data-generating models and hence a standard supervised learning problem (regression or classification). This allows us to adopt any partition-based methods to learn the subgroup from data. We instantiate the approach with CART, arguably one of the most widely used tree-based methods, to learn the subgroup with maximum treatment effect. Finally, on a large collection of synthetic and semi-synthetic datasets, we compare our method against a wide range of baselines and find that our approach, which avoids such causal heuristics, more accurately identifies subgroups with maximum treatment effect. Our source code is available at https://github.com/ylincen/causal-subgroup.
- 23 September 2025
- Published by Springer Nature
- p. 56-74
Snor: Simpler Descriptions Through Overlapping Patterns
- 23 September 2025
- Published by Springer Nature
- p. 56-74
Snor: Simpler Descriptions Through Overlapping Patterns
- Open Access
- 5 July 2025
- Published by Springer Nature in Data Mining and Knowledge Discovery
- Vol. 39 (5) , 1-43
Towards automated self-supervised learning for truly unsupervised graph anomaly detection
Self-supervised learning (SSL) is an emerging paradigm that exploits supervisory signals generated from the data itself, and many recent studies have leveraged SSL to conduct graph anomaly detection. However, we empirically found that three important factors can substantially impact detection performance across datasets: (1) the specific SSL strategy employed; (2) the tuning of the strategy’s hyperparameters; and (3) the allocation of combination weights when using multiple strategies. Most SSL-based graph anomaly detection methods circumvent these issues by arbitrarily or selectively (i.e., guided by label information) choosing SSL strategies, hyperparameter settings, and combination weights. While an arbitrary choice may lead to subpar performance, using label information in an unsupervised setting is label information leakage and leads to severe overestimation of a method’s performance. Leakage has been criticized as “one of the top ten data mining mistakes", yet many recent studies on SSL-based graph anomaly detection have been using label information to select hyperparameters. To mitigate this issue, we propose to use an internal evaluation strategy (with theoretical analysis) to select hyperparameters in SSL for unsupervised anomaly detection. We perform extensive experiments using 10 recent SSL-based graph anomaly detection algorithms on various benchmark datasets, demonstrating both the prior issues with hyperparameter selection and the effectiveness of our proposed strategy.
- 5 July 2025
- Published by Springer Nature in Data Mining and Knowledge Discovery
- Vol. 39 (5) , 1-43
Open AccessTowards automated self-supervised learning for truly unsupervised graph anomaly detection
Self-supervised learning (SSL) is an emerging paradigm that exploits supervisory signals generated from the data itself, and many recent studies have leveraged SSL to conduct graph anomaly detection. However, we empirically found that three important factors can substantially impact detection performance across datasets: (1) the specific SSL strategy employed; (2) the tuning of the strategy’s hyperparameters; and (3) the allocation of combination weights when using multiple strategies. Most SSL-based graph anomaly detection methods circumvent these issues by arbitrarily or selectively (i.e., guided by label information) choosing SSL strategies, hyperparameter settings, and combination weights. While an arbitrary choice may lead to subpar performance, using label information in an unsupervised setting is label information leakage and leads to severe overestimation of a method’s performance. Leakage has been criticized as “one of the top ten data mining mistakes", yet many recent studies on SSL-based graph anomaly detection have been using label information to select hyperparameters. To mitigate this issue, we propose to use an internal evaluation strategy (with theoretical analysis) to select hyperparameters in SSL for unsupervised anomaly detection. We perform extensive experiments using 10 recent SSL-based graph anomaly detection algorithms on various benchmark datasets, demonstrating both the prior issues with hyperparameter selection and the effectiveness of our proposed strategy.
- Open Access
- 26 June 2025
- Published by Elsevier in Artificial Intelligence in Medicine
- Vol. 167, 103200
Discovering multiple antibiotic resistance phenotypes using diverse top-k subgroup list discovery
- 26 June 2025
- Published by Elsevier in Artificial Intelligence in Medicine
- Vol. 167, 103200
Open AccessDiscovering multiple antibiotic resistance phenotypes using diverse top-k subgroup list discovery
- Open Access
- 16 January 2025
- Published by Open Exploration Publishing
- Vol. 3, 101137
Using consumer wearables to estimate physical activity of nursing home residents with dementia
Aim: Physical activity of nursing home residents can be assessed with tools such as questionnaires and standardized fitness tests. For residents with dementia, however, those tools can be cognitively challenging and difficult to administer. Consumer wearables could potentially aid as an affordable tool for ubiquitous assessment. Methods: In this pilot study with 16 participants, we explored how measurements with an off-the-shelf wearable relate to structured observations of physical activity. We collected both processed and raw tri-axial accelerometer data from Samsung wrist-worn fitness trackers. To anchor those data in the free-living environment, we compared the measurements with the physical activity scale of the Medlo behavioral observation scheme. Results: We showed that consumer wearables are a valid tool for long-term data collection in this vulnerable patient population. Conclusions: Regarding the movement intensity, the data collected by fitness trackers is overall in accordance with the data collected with the observational tool. Regarding the type of movement, we concluded that the automatic activity classification on the wearables is not yet ready for use with a mostly sedentary patient population.
- 16 January 2025
- Published by Open Exploration Publishing
- Vol. 3, 101137
Open AccessUsing consumer wearables to estimate physical activity of nursing home residents with dementia
Aim: Physical activity of nursing home residents can be assessed with tools such as questionnaires and standardized fitness tests. For residents with dementia, however, those tools can be cognitively challenging and difficult to administer. Consumer wearables could potentially aid as an affordable tool for ubiquitous assessment. Methods: In this pilot study with 16 participants, we explored how measurements with an off-the-shelf wearable relate to structured observations of physical activity. We collected both processed and raw tri-axial accelerometer data from Samsung wrist-worn fitness trackers. To anchor those data in the free-living environment, we compared the measurements with the physical activity scale of the Medlo behavioral observation scheme. Results: We showed that consumer wearables are a valid tool for long-term data collection in this vulnerable patient population. Conclusions: Regarding the movement intensity, the data collected by fitness trackers is overall in accordance with the data collected with the observational tool. Regarding the type of movement, we concluded that the automatic activity classification on the wearables is not yet ready for use with a mostly sedentary patient population.
- Open Access
Conditional Density Estimation with Histogram Trees
Conditional density estimation (CDE) goes beyond regression by modeling the full conditional distribution, providing a richer understanding of the data than just the conditional mean in regression. This makes CDE particularly useful in critical application domains. However, interpretable CDE methods are understudied. Current methods typically employ kernel-based approaches, using kernel functions directly for kernel density estimation or as basis functions in linear models. In contrast, despite their conceptual simplicity and visualization suitability, tree-based methods -- which are arguably more comprehensible -- have been largely overlooked for CDE tasks. Thus, we propose the Conditional Density Tree (CDTree), a fully non-parametric model consisting of a decision tree in which each leaf is formed by a histogram model. Specifically, we formalize the problem of learning a CDTree using the minimum description length (MDL) principle, which eliminates the need for tuning the hyperparameter for regularization. Next, we propose an iterative algorithm that, although greedily, searches the optimal histogram for every possible node split. Our experiments demonstrate that, in comparison to existing interpretable CDE methods, CDTrees are both more accurate (as measured by the log-loss) and more robust against irrelevant features. Further, our approach leads to smaller tree sizes than existing tree-based models, which benefits interpretability.
Open AccessConditional Density Estimation with Histogram Trees
Conditional density estimation (CDE) goes beyond regression by modeling the full conditional distribution, providing a richer understanding of the data than just the conditional mean in regression. This makes CDE particularly useful in critical application domains. However, interpretable CDE methods are understudied. Current methods typically employ kernel-based approaches, using kernel functions directly for kernel density estimation or as basis functions in linear models. In contrast, despite their conceptual simplicity and visualization suitability, tree-based methods -- which are arguably more comprehensible -- have been largely overlooked for CDE tasks. Thus, we propose the Conditional Density Tree (CDTree), a fully non-parametric model consisting of a decision tree in which each leaf is formed by a histogram model. Specifically, we formalize the problem of learning a CDTree using the minimum description length (MDL) principle, which eliminates the need for tuning the hyperparameter for regularization. Next, we propose an iterative algorithm that, although greedily, searches the optimal histogram for every possible node split. Our experiments demonstrate that, in comparison to existing interpretable CDE methods, CDTrees are both more accurate (as measured by the log-loss) and more robust against irrelevant features. Further, our approach leads to smaller tree sizes than existing tree-based models, which benefits interpretability.
- Open Access
- 20 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 36 (12) , 7839-7850
Cross-Domain Graph Level Anomaly Detection
Existing graph level anomaly detection methods are predominantly unsupervised due to high costs for obtaining labels, yielding sub-optimal detection accuracy when compared to supervised methods. Moreover, they heavily rely on the assumption that the training data exclusively consists of normal graphs. Hence, even the presence of a few anomalous graphs can lead to substantial performance degradation. To alleviate these problems, we propose a cross-domain graph level anomaly detection method, aiming to identify anomalous graphs from a set of unlabeled graphs (target domain) by using easily accessible normal graphs from a different but related domain (source domain). Our method consists of four components: a feature extractor that preserves semantic and topological information of individual graphs while incorporating the distance between different graphs; an adversarial domain classifier to make graph level representations domain-invariant; a one-class classifier to exploit label information in the source domain; and a class aligner to align classes from both domains based on pseudolabels. Experiments on seven benchmark datasets show that the proposed method largely outperforms state-of-the-art methods
- 20 September 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Knowledge and Data Engineering
- Vol. 36 (12) , 7839-7850
Open AccessCross-Domain Graph Level Anomaly Detection
Existing graph level anomaly detection methods are predominantly unsupervised due to high costs for obtaining labels, yielding sub-optimal detection accuracy when compared to supervised methods. Moreover, they heavily rely on the assumption that the training data exclusively consists of normal graphs. Hence, even the presence of a few anomalous graphs can lead to substantial performance degradation. To alleviate these problems, we propose a cross-domain graph level anomaly detection method, aiming to identify anomalous graphs from a set of unlabeled graphs (target domain) by using easily accessible normal graphs from a different but related domain (source domain). Our method consists of four components: a feature extractor that preserves semantic and topological information of individual graphs while incorporating the distance between different graphs; an adversarial domain classifier to make graph level representations domain-invariant; a one-class classifier to exploit label information in the source domain; and a class aligner to align classes from both domains based on pseudolabels. Experiments on seven benchmark datasets show that the proposed method largely outperforms state-of-the-art methods
- Open Access
- 14 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 306-307
Graph Neural Networks based Log Anomaly Detection and Explanation
Event logs are widely used to record the status of high-tech systems, making log anomaly detection important for monitoring those systems. We propose a graph-based method for unsupervised log anomaly detection, dubbed Logs2Graphs, which first converts event logs into attributed, directed, and weighted graphs, and then leverages graph neural networks to perform graph-level anomaly detection. Specifically, we introduce OCDiGCN, a novel graph neural network model for detecting graph-level anomalies in a collection of attributed, directed, and weighted graphs. By coupling the graph representation and anomaly detection steps, OCDiGCN can learn a representation that is especially suited for anomaly detection, resulting in a high detection accuracy. For each detected anomaly, we provide a subset of nodes that are crucial in OCDiGCN's predictions, offering useful insights for root cause diagnosis. Experiments on five benchmark datasets show that Logs2Graphs matches or exceeds current top log anomaly detection methods on simple datasets and largely outperforms them on complex ones.
- 14 April 2024
- Published by Association for Computing Machinery (ACM)
- p. 306-307
Open AccessGraph Neural Networks based Log Anomaly Detection and Explanation
Event logs are widely used to record the status of high-tech systems, making log anomaly detection important for monitoring those systems. We propose a graph-based method for unsupervised log anomaly detection, dubbed Logs2Graphs, which first converts event logs into attributed, directed, and weighted graphs, and then leverages graph neural networks to perform graph-level anomaly detection. Specifically, we introduce OCDiGCN, a novel graph neural network model for detecting graph-level anomalies in a collection of attributed, directed, and weighted graphs. By coupling the graph representation and anomaly detection steps, OCDiGCN can learn a representation that is especially suited for anomaly detection, resulting in a high detection accuracy. For each detected anomaly, we provide a subset of nodes that are crucial in OCDiGCN's predictions, offering useful insights for root cause diagnosis. Experiments on five benchmark datasets show that Logs2Graphs matches or exceeds current top log anomaly detection methods on simple datasets and largely outperforms them on complex ones.