Estimating Dynamic Graphical Models from Multivariate Time-Series Data: Recent Methods and Results

Optimising dynamic graphical models for video content analysis

Multi-Camera Multiple Object Tracking (MC-MOT) is a significant computer vision problem due to its emerging applicability in several real-world applications. Despite a large number of existing works, solving the data association problem in any MC-MOT pipeline is arguably one of the most challenging tasks. Developing a robust MC-MOT system, however, is still highly challenging due to many practical issues such as inconsistent lighting conditions, varying object movement patterns, or the trajectory occlusions of the objects between the cameras. To address these problems, this work, therefore, proposes a new Dynamic Graph Model with Link Prediction (DyGLIP) approach <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup> to solve the data association task. Compared to existing methods, our new model offers several advantages, including better feature representations and the ability to recover from lost tracks during camera transitions. Moreover, our model works gracefully regardless of the overlapping ratios between the cameras. Experimental results show that we out-perform existing MC-MOT algorithms by a large margin on several practical datasets. Notably, our model works favor-ably on online settings but can be extended to an incremental approach for large-scale datasets.

In this work, a dynamic graphical model of operating modes of intelligent and autonomous multi inputs multi outputs (MIMO) vehicles is presented. This study shows the feasibility of improving the traffic management and decision inside confined space according to the operating situation of each involved vehicle. The proposed dynamic graphical model depends on the on-line monitoring outputs of the vehicles' actuators. These outputs correspond to the dynamic model based residuals of the actuators. Co-simulation using experimental data show the interest of the developed model.

Dynamic networks require effective methods of monitoring and surveillance in order to respond promptly to unusual disturbances. In many applications, it is of interest to identify anomalous behavior within a dynamic interacting system. Such anomalous interactions are reflected by structural changes in the network representation of the system. In this paper, a dynamic random graph model is proposed that takes into account the past activities of the individuals in the social network and also represents temporal dependency of the network. The model parameters are appearance and disappearance probabilities of an edge which are estimated using a maximum likelihood approach. A generalization of a single path‐dependent likelihood ratio test is employed to detect changes in the parameters of the proposed model. Through monitoring the estimated parameters, one can effectively detect structural changes in a temporal‐dependent network. The proposed model is employed to describe the behavior of a real network, and its parameters are monitored via dependent likelihood ratio test and multivariate exponentially weighted moving average control chart. Results indicate that the proposed dynamic random graph model is a reliable mean to modeling and detecting changes in temporally dependent networks.

Intent detection and slot filling are two main tasks for building a spoken language understanding (SLU) system. Since the two tasks are closely related, the joint models for the two tasks always outperform the pipeline models in SLU. However, most joint models directly incorporate multiple intent information for each token, which introduces intent noise into the sentence semantics, causing a decrease in the performance of the joint model. In this paper, we propose a Dynamic Graph Model (DGM) for joint multiple intent detection and slot filling, in which we adopt a sentence-level intent-slot interactive graph to model the correlation between the intents and slot. Besides, we design a novel method of constructing the graph, which can dynamically update the interactive graph and further alleviate the error propagation. Experimental results on several multi-intent and single-intent datasets show that our model not only achieves the state-of-the-art (SOTA) performance but also boosts the speed by three to six times over the SOTA model.

Continuous-time dynamic graph modeling is a crucial task for many real-world applications, such as financial risk management and fraud detection. Though existing dynamic graph modeling methods have achieved satisfactory results, they still suffer from three key limitations, hindering their scalability and further applicability. i) Indiscriminate updating. For incoming edges, existing methods would indiscriminately deal with them, which may lead to more time consumption and unexpected noisy information. ii) Ineffective node-wise long-term modeling. They heavily rely on recurrent neural networks (RNNs) as a backbone, which has been demonstrated to be incapable of fully capturing node-wise long-term dependencies in event sequences. iii) Neglect of re-occurrence patterns. Dynamic graphs involve the repeated occurrence of neighbors that indicates their importance, which is disappointedly neglected by existing methods.

In today’s networked systems a massive amount of data is produced every day. These data can be modelled using graphs, where the nodes typically correspond to users or devices and the edges to the connections between them. Almost all networks change over time, with new nodes and edges appearing or disappearing as the system matures. Therefore dynamic graph models are more adequate to analyse such networks than static graphs, and appropriate tools need to be implemented to protect them. In this paper we obtain an edge differentially private version of dynamic stochastic block model. We show experimentally that the trends in the dynamic stochastic block model obtained from the original data are well preserved with the additional privacy guarantees.

The recommender systems can gain the needs and interests of users by analyzing the user history data and then help the users making decisions on appropriate choices in E-commerce. However, with the increasing of data volume and the popularization of information network, the participation of users in E-commerce activities is growing deeply. How to analyze the user preferences and make a user-centered efficient recommendation is an urgent problem to be further researched. In this paper, we first propose the user-centered recommendation based on dynamic graph model to express the user preferences and gain the user preference vectors for recommendation. Then, after gaining the user preferences vectors, we propose the user clustering algorithm using US-ELM to cluster the users into different clusters. Last, we provide two recommendation algorithms, which can present top-k recommendation, respectively the group recommendation and personal recommendation. With the extensive experiments, our recommendation algorithms can effectively express the user preferences and reach a good performance.

Moment-based estimation of smooth transition regression models with endogenous variables

Multivariate time series data often have a very high dimensionality. Classifying such high dimensional data poses a challenge because a vast number of features can be extracted. Furthermore, the meaning of the normally intuitive term similar needs to be precisely defined. Representing the time series data effectively is an essential task for decision-making activities such as prediction, clustering and classification. In this paper we propose a feature-based classification approach to classify real-world multivariate time series generated by drilling rig sensors in the oil and gas industry. Our approach encompasses two main phases: representation and classification. For the representation phase, we propose a novel representation of time series which combines trend-based and value-based approximations (we abbreviate it as TVA). It produces a compact representation of the time series which consists of symbolic strings that represent the trends and the values of each variable in the series. The TVA representation improves both the accuracy and the running time of the classification process by extracting a set of informative features suitable for common classifiers. For the classification phase, we propose a memory-based classifier which takes into account the antecedent results of the classification process. The inputs of the proposed classifier are the TVA features computed from the current segment, as well as the predicted class of the previous segment. Our experimental results on real-world multivariate time series show that our approach enables highly accurate and fast classification of multivariate time series.

We introduce a method for expressing word lattices within a dynamic graphical model. We describe a variety of choices for doing this, including a technique to relax the time information associated with lattice nodes in a way that trades off hypothesis expansion with presumed segmentation boundary accuracy. Our approach uses a set of time-inhomogeneous and algorithmically expressed conditional probability tables to encode the lattice. The approach was implemented as part of the graphical model toolkit, and word error rate improvements on the Switchboard corpus indicate that our technique is a viable means to incorporate large state space speech recognition systems into a graphical model.

Operational and financial performance of Italian airport companies: A dynamic graphical model

We introduced a novel methodology Multivariate Degree Distribution to Dynamic Graph (MDD2DG) with Inter-channel Relevance Attention (IRA) mechanism to analyze multi-channel Electrocardiogram (ECG) signals and explore signal connections across different channels. Our methodology comprises three main steps. First, multi-channel cardiac signals are transformed into multi-channel visual graphs to extract crucial degree distribution features. Then, degree distributions are mapped into dynamic graphs using a neural network with an IRA mechanism. After that, critical features are extracted within dynamic graphs utilizing a Graph Convolutional Neural Networks (GCNNs), and classification is subsequently performed using a multilayer perceptron. In this model, a method of multi-scale position embedding was introduced, which significantly enhanced the processing efficiency of the model by providing a simpler yet sufficiently effective feature representation. Compared to traditional complex network methods, our approach replaces fixed formula-calculated features with dynamic graph models, resulting in improved recognition accuracy. In the experiments, we achieved an impressive 99.94% classification accuracy for distinguishing ECG signals from the five distinct locations (AMI, ASMI, ALMI, IMI and ILMI) with myocardial infarction (MI) as well as those of the healthy controls (HC). This work contributes to the analysis of complex physiological signals in the field of multi-channel ECG sequence, and provides a robust approach with promising implications for improving clinical medicine and the early detection of cardiac diseases.

Mental disorders often emerge during adolescence and have been associated with age-related differences in connection strengths of brain networks (static functional connectivity), manifesting in non-typical trajectories of brain development. However, little is known about the direction of information flow (directed functional connectivity) in this period of functional brain progression. We employed dynamic graphical models (DGM) to estimate directed functional connectivity from resting state functional magnetic resonance imaging data on 1143 participants, aged 6 to 17 years from the healthy brain network (HBN) sample. We tested for effects of age, sex, cognitive abilities and psychopathology on estimates of direction flow. Across participants, we show a pattern of reciprocal information flow between visual-medial and visual-lateral connections, in line with findings in adults. Investigating directed connectivity patterns between networks, we observed a positive association for age and direction flow from the cerebellar to the auditory network, and for the auditory to the sensorimotor network. Further, higher cognitive abilities were linked to lower information flow from the visual occipital to the default mode network. Additionally, examining the degree networks overall send and receive information to each other, we identified age-related effects implicating the right frontoparietal and sensorimotor network. However, we did not find any associations with psychopathology. Our results suggest that the directed functional connectivity of large-scale resting-state brain networks is sensitive to age and cognition during adolescence, warranting further studies that may explore directed relationships at rest and trajectories in more fine-grained network parcellations and in different populations.

Prediction of anomaly in time-varying datasets is at the core of many real-world applications. For example, prediction of anomalous events in different time-varying business metrics and proactively mitigating any underlying problem in the system is of utmost importance to any online business to evade any kind of disruption in smooth customer experience. A large body of literature exists which adopted different techniques for anomaly prediction in multi-variate time-series data. Recent advancements in deep-learning have resulted in multiple literature proposing different variants of graph neural network (GNN) for anomaly prediction in multi-variate time-series data. However, one major gap area that remained in this domain is the sensitivity of such networks to inherent noise in the data. This is where we make our contribution by proposing a graph neural network with dynamic attention which not only offers more robustness of the network to noise but also makes the network more trainable and accurate in prediction in the presence of noise. We carried out extensive experimentation against both publicly available datasets and real-world industrial datasets. The outcome of the experimentation is consistently in support of the superiority of our proposed network over the state-of-the-art GNN network for anomaly prediction.