In this study, we propose a novel neuro-symbolic approach to deal with the inherent ambiguity in image scene classification, combining the usage of pre-trained deep learning (DL) models with concepts from modal logic and evidence theory. The DL models are used to detect objects and estimate their depth in a set of labeled images. The obtained outputs are employed to form a dataset of instances characterizing the possible classes. Subsequently, a multi-valued mapping is defined between the data instances and the considered images resulting into each image being represented by the set of instances associated with it. The obtained mapping is utilized to infer necessity and possibility conditions of each class, or equivalently its upper (plausibility) and lower (belief) probabilities. Based on these interval evaluations, a rule-based and a score-based classifiers are built. The overall method is explainable and directly interpretable, robust to data scarcity and data imbalance. The presented framework is studied and evaluated on an abandoned bag detection use case.

In the rapidly evolving landscape of mobility modelling, the application of deep learning approaches introduces both opportunities and challenges. Such approaches, while powerful, yield opaque models that lack interpretability and adaptability to diverse traffic contexts. Addressing those challenges, a finite set of humanly-interpretable traffic states is exploited here for the purpose of facilitating the annotation of mobility data with meaningful labels such as congestion, free-flow, traffic build-up, etc. Such annotation unlocks a range of opportunities to integrate multiple complementary approaches for modelling state transition behaviour. Concretely, a novel hybrid modelling framework is introduced in this article leveraging multiple data representations (temporal, time-frequency and symbolic) with the aim to forecast traffic progression in terms of humanly-explicable state transitions. Three distinct modelling paradigms are subsequently explored: neural, neural-to-symbolic, and symbolic-to-neural, by demonstrating their potential to capture and forecast traffic dynamics on real-world mobility data. While the fully neural approach is undoubtedly the most accurate one, the two neuro-symbolic approaches offer a better trade-off between accuracy on one side and interpretability, probability calibration, and computational efficiency, on the other. This work illustrates the importance of tailored data representations in understanding and predicting complex mobility behaviour, highlighting the benefits of hybrid approaches in achieving interpretability and efficiency in traffic data analysis.

In this study, we propose a data-driven survival risk analysis approach in support of predictive maintenance management of a large portfolio of industrial assets. The concrete use case considered is a large portfolio of industrial vehicles (trucks). However, the approach is generic (i.e., asset-type agnostic) in nature and can be applied in different industrial contexts. It is able to employ different data sources in the risk analysis workflow, e.g., time series operation data collected via a multitude of sensor measurements combined with tabular data recording the technical specifications of the assets (vehicles). Subsequently, several different risk assessment strategies can be considered: 1) operation-related risk at each time step for any asset computed on the operation data across the whole portfolio; 2) the failure predisposition of each asset determined by its technical specification; 3) hybrid risk analysis, which innovatively combines the different data types to estimate overall risk at any time in the future for any asset. Our validation, conducted on real-world data, demonstrates that the hybrid approach provides a realistic temporal risk assessment during vehicle operation that also reflects adequately the inherent (contextual) risk predisposition of the vehicle due its technical specification. The proposed approach derives diverse survival risk estimations, which are interpretable by design and in this way facilitate both prognostic health monitoring and root cause analysis of the factors impacting vehicles’ risk of failure.

Wind turbines operate under different regimes depending on the contextual conditions, such as meteorology, operational constraints, or grid loads. Knowing when a turbine changes operational regime (its current operating mode) is crucial for a proper assessment of its performance, as only then can one compare it to the expected performance. However, often operators overseeing the turbines do not have access to the operating mode (OM) a turbine is in, leading to incomplete or inaccurate evaluation of its current capacity. We have devised an approach to leverage the knowledge across different wind farms aiming at reliably transferring, using domain adaptation, an OM detection model from a source turbine, for which OM labels are available, to a target turbine, for which they are not. This is achieved by an active learning process, enabling the adaptation to take place with expert supervision, while minimizing the required effort. Furthermore, an extension based on density-based clustering allows to post-process the outcome of the active learning to further characterize and discriminate between multiple non-primary modes of the turbine. Our approach is validated on a real-life dataset of onshore wind turbines, and we demonstrate that our active domain adaptation method achieves optimal OM detection performance.

Distributed systems, characterized by their inherent heterogeneity, frequently encounter challenges in managing and interpreting a multitude of events. In this study, we advocate for a standardized methodology designed to address these issues in diverse distributed environments. The proposed approach focuses on identifying spatiotemporal contextual factors and understanding inter-and intra-asset dependencies. Our methodology utilizes the Markov clustering algorithm to enable a systematic exploration of event data across different contexts. We illustrate the versatility of this approach by applying it to the specific domain of telecommunication network alarms. The results demonstrate how the methodology efficiently clusters and manages alarm data within complex distributed systems, offering a valuable tool for general event management in diverse environments.

This paper proposes a novel federated learning methodology, called FedRepo, that copes with concept drift issues in a statistically heterogeneous distributed learning environment. The proposed horizontal federated learning methodology, based on random forest (RF), can be used for collaborative training and maintenance of a dynamic repository of federated RF models, each one customized to a group of clients/devices. The clients are grouped together if their performance patterns with respect to the global RF model are similar. The performance of the customized RF global models is continuously monitored during the inference phase and the repository is accordingly adapted to mitigate the detected concept drift. The proposed methodology is studied and evaluated against an electricity consumption forecasting use case. The evaluation results demonstrate clearly that the proposed methodology is able to deal with concept drift issues in an efficient and adequate fashion without compromising the overall performance of the distributed environment.

In this study, we propose a new multi-view stream clustering approach, called MV Split-Merge Clustering. The proposed approach is an extension of an existing split-merge evolutionary clustering algorithm (entitled Split-Merge Clustering) to multi-view data applications. The extended version can be used to integrate data from multiple views in a streaming manner and discover cluster structure for each data chunk. The MV Split-Merge Clustering can be applied for grouping distinct chunks of multi-view streaming data so that a global integrated clustering model is built on each data chunk. At each time window, an updated clustering solution (local model) is initially produced on each view of the current data chunk by applying the Split-Merge Clustering algorithm. Formal Concept Analysis is then used in order to integrate information from the multiple views (local clustering models) and generate a global model (formal concept lattice) that reveals the correlations among the clusters of the local models. The proposed MV Split-Merge Clustering has been initially evaluated on a publicly available data set. Our results show that the approach is able to identify a clustering structure and relationships among the different views comparable to those produced in a batch scenario.

In this article we propose a bipartite correlation clustering technique that can be used to adapt the existing clustering solution to a clustering of newly collected data elements. The proposed tec ...

The problem addressed in this article concerns the development of evolutionary clustering techniques that can be applied to adapt the existing clustering solution to a clustering of newly collected ...

Nowadays, more and more industrial assets are continuously monitored and generate vast amount of events and sensor data. It provides an excellent opportunity for understanding the asset behaviour that is currently underexplored due to several challenges: extremely heterogeneous data sources, overwhelming data volume, textual aspect of event logs and complex relational dependencies between events. We have addressed this problem by developing two methodologies: 1) A methodology to detect the relevant events while taking into account the relations between them 2) A methodology (built on top of the first one) to build performance profiles taking into account multiple data sources (events and sensor data). We have validated the methodologies in the specific photovoltaic (PV) domain.

Solar plants typically consist of several thousands of passive photovoltaic modules that are connected via thousands of string boxes to hundreds of inverters. In addition, a solar plant has meteo-sensors, power meters and control switches. All these components continuously generate data that is collected by monitoring systems or SCADA systems on-site. From there onwards, this data is pushed to remote analysis servers. The optimal exploitation of this data is hampered by a lack of harmonisation and standardisation in the photovoltaic domain. The data-generating components originate from several different manufacturers, models and versions, and their output is thus not easily commensurable. Crucial for this paper is the fact that conceptually identical failure events are not logged with the same identifying labels. Therefore, every analysis of monitoring system data coming from photovoltaic plants needs an initial integration step to resolve this labeling issue. Our proposal is to facilitate the integration with semantic modelling by means of creating a photovoltaic event ontology with an SWRL reasoning layer.

BACKGROUND: Presently, with the increasing number and complexity of available gene expression datasets, the combination of data from multiple microarray studies addressing a similar biological question is gaining importance. The analysis and integration of multiple datasets are expected to yield more reliable and robust results since they are based on a larger number of samples and the effects of the individual study-specific biases are diminished. This is supported by recent studies suggesting that important biological signals are often preserved or enhanced by multiple experiments. An approach to combining data from different experiments is the aggregation of their clusterings into a consensus or representative clustering solution which increases the confidence in the common features of all the datasets and reveals the important differences among them. RESULTS: We propose a novel generic consensus clustering technique that applies Formal Concept Analysis (FCA) approach for the consolidation and analysis of clustering solutions derived from several microarray datasets. These datasets are initially divided into groups of related experiments with respect to a predefined criterion. Subsequently, a consensus clustering algorithm is applied to each group resulting in a clustering solution per group.These solutions are pooled together and further analysed by employing FCA which allows extracting valuable insights from the data and generating a gene partition over all the experiments. In order to validate the FCA-enhanced approach two consensus clustering algorithms are adapted to incorporate the FCA analysis. Their performance is evaluated on gene expression data from multi-experiment study examining the global cell-cycle control of fission yeast. The FCA results derived from both methods demonstrate that, although both algorithms optimize different clustering characteristics, FCA is able to overcome and diminish these differences and preserve some relevant biological signals. CONCLUSIONS: The proposed FCA-enhanced consensus clustering technique is a general approach to the combination of clustering algorithms with FCA for deriving clustering solutions from multiple gene expression matrices. The experimental results presented herein demonstrate that it is a robust data integration technique able to produce good quality clustering solution that is representative for the whole set of expression matrices.

In this paper we present Emergencia, a software tool prototype to support the coordinator of an emergency (e.g. at a large industrial site) in the process of taking the appropriate decisions to solve the emergency situation with minimal impact on safety and devastation. Such a coordinator usually performs this task with the help of an emergency plan, a printed document describing relevant procedures to follow during an emergency. Emergency plans of large sites can easily contain hundreds of pages and might be difficult to use when decisions need to be taken quickly. Instead of browsing such a document in search for the relevant procedures, Emergencia guides the coordinator through this document and pro-actively suggests appropriate procedures to follow and pertinent decisions and measures to take, by considering the current status of the emergency situation.

MOTIVATION: The validity of periodic cell cycle regulation studies in plants is seriously compromised by the relatively poor quality of cell synchrony that is achieved for plant suspension cultures in comparison to yeast and mammals. The present state-of-the-art plant synchronization techniques cannot offer a complete cell cycle coverage and moreover a considerable loss of cell synchrony may occur toward the end of the sampling. One possible solution is to consider combining multiple datasets, produced by different synchronization techniques and thus covering different phases of the cell cycle, in order to arrive at a better cell cycle coverage. RESULTS: We propose a method that enables pasting expression profiles from different plant cell synchronization experiments and results in an expression curve that spans more than one cell cycle. The optimal pasting overlap is determined via a dynamic time warping alignment. Consequently, the different expression time series are merged together by aggregating the corresponding expression values lying within the overlap area. We demonstrate that the periodic analysis of the merged expression profiles produces more reliable p-values for periodicity. Subsequent Gene Ontology analysis of the results confirms that merging synchronization experiments is a more robust strategy for the selection of potentially periodic genes. Additional validation of the proposed algorithm on yeast data is also presented. AVAILABILITY: Results, benchmark sets and scripts are freely available at our website: http://www.psb.ugent.be/cbd/publications.php

UNLABELLED: An application tool for alignment, template matching and visualization of gene expression time series is presented. The core algorithm is based on dynamic time warping techniques used in the speech recognition field. These techniques allow for non-linear (elastic) alignment of temporal sequences of feature vectors and consequently enable detection of similar shapes with different phases. AVAILABILITY: The Java program, examples and a tutorial are available at http://www.psb.ugent.be/cbd/papers/gentxwarper/

Three baseline isolated word rejection strategies are initially proposed and their rejection performance is evaluated on two different types of out-of-vocabulary (OOV) utterances: OOV similar in nature to the in-vocabulary (IV) ones versus OOV consisting of non-speech events such as coughs, clicks, smacks, etc. A general OOV model, referred to as garbage, is added in parallel to the IV models and then a confidence measure, based on the contrast of the first best, hypothesis score with the garbage score, is used as an IV-OOV classifier. The discriminative power of some other confidence measures, utilising the distance between the first two hypotheses in the N-best list, is also investigated. Further, a considerable improvement (up to 34%) of the baseline classification error rate is achieved when the garbage model is supplied with word transition penalties.

Modal logic interpretations of plausibility and belief measures are developed based on the observation that the accessibility relation in a model of modal logic, regarded as a multivalued mapping, induces a plausibility measure and a belief measure on the set of possible worlds.

We show that a (fuzzy) multivalued mapping carries a possibility and necessity measure defined over (fuzzy) subsets of a first universe into a system of upper and lower possibilities and necessities defined over (fuzzy) subsets of a second universe. Upper possibilities (resp. lower necessities) form again a possibility measure (resp. necessity measure), while lower possibilities and upper necessities both form a confidence measure. The approach presented is based on possibilistic conditioning, in particular on the definitions of conditional possibilities and necessities in a general framework based on triangular norms and conorms. In case of a multivalued mapping, upper and lower possibilities and necessities can be expressed equivalently in terms of conditional possibilities and necessities of lower and upper inverse images under the given multivalued mapping, or in terms of a basic possibility assignment and a basic necessity assignment. In case of a fuzzy multivalued mapping, such equivalent expressions cannot be established in general. Upper and lower possibilities and necessities can then be introduced in two alternative ways. For normalized fuzzy multivalued mappings or crisp multivalued mappings, interesting relationships are obtained.