Search

This paper introduces a novel anomaly detection approach tailored for time series data with exclusive reliance on normal events during training. Our key innovation lies in the application of kernel-density estimation (KDE) to scrutinize reconstruction errors, providing an empirically derived probability distribution for normal events post-reconstruction. This non-parametric density estimation technique offers a nuanced understanding of anomaly detection, differentiating it from prevalent threshold-based mechanisms in existing methodologies. In post-training, events are encoded, decoded, and evaluated against the estimated density, providing a comprehensive notion of normality. In addition, we propose a data augmentation strategy involving variational autoencoder-generated events and a smoothing step for enhanced model robustness. The significance of our autoencoder-based approach is evident in its capacity to learn normal representation without prior anomaly knowledge. Through the KDE step on reconstruction errors, our method addresses the versatility of anomalies, departing from assumptions tied to larger reconstruction errors for anomalous events. Our proposed likelihood measure then distinguishes normal from anomalous events, providing a concise yet comprehensive anomaly detection solution. The extensive experimental results support the feasibility of our proposed method, yielding significantly improved classification performance by nearly 10% on the UCR benchmark data.

Accurately measuring vehicle mileage is pivotal in precise CO2 emission calculations and the development of reliable emission models. Nonetheless, mileage data gathered from surveys relying on self-estimation, garage reports, and other estimation-based sources often yield rough approximations that substantially deviate from the actual mileage. To tackle this issue, we present a comprehensive framework aimed at bolstering the accuracy of CO2 emission models. This paper harnesses two innovative techniques: the deep learning semi-supervised fuzzy C-means (SSFCM) and polynomial classifier models. By leveraging these sophisticated mathematical techniques, we achieve successful classification of passenger vehicles, enabling more precise evaluations of average mileage. Real data shows that vehicles in Switzerland considerably exceed the estimated mileage in the years following the first registration of the vehicle. The difference lies in the covered mileage after vehicles reach five years of age. Our framework supports segment-based analysis for assessing average mileage and enhancing emission models for better understanding of vehicle-related environmental impact.

Quantum computing promises to solve difficult optimization problems in chemistry, physics and mathematics more efficiently than classical computers. However, it requires fault-tolerant quantum computers with millions of qubits; a technological challenge still not mastered by engineers. To lower the barrier, hybrid algorithms combining classical and quantum computers are used, where quantum computing is only used for those parts of computation that cannot be solved efficiently otherwise. In this paper, we tackle the multiple query optimization problem (MQO), an important NP-hard problem in database research. We present an implementation based on a scheme called quantum approximate optimization algorithm to solve the MQO on a gate-based quantum computer. We perform a detailed experimental evaluation of our implementation and compare its performance against a competing approach that employs a quantum annealer – another type of quantum computer. Our implementation shows a qubit efficiency of close to 99%, which is almost a factor of 2 higher than the state-of-the-art implementation. We emphasize that the problems we can solve with current gate-based quantum technology are fairly small and might not seem practical yet compared to state-of-the-art classical query optimizers. However, our experiments on using a hybrid approach of classical and quantum computing show that our implementation scales favourably with larger problem sizes. Hence, we conclude that our approach shows promising results for near-term quantum computers and thus sets the stage for a challenging avenue of novel database research.

Automated chart analysis has vast potential to improve the accessibility of charts for a wider audience, e.g., people with visual impairments or other disabilities, by generating captions for chart images that can quickly convey the information being represented. Additionally, it can improve the performance of automatic document analysis systems, by enabling them to extract valuable information from the documents with graphical/visual scientific content. Although recent advancements in modality translation and multi-modal learning have led to the development of more or less successful image captioning and visual question-answering methods, but most of them have been designed for general images, and cannot be successfully applied to specific areas such as medical images or scientific charts and graphs. Therefore, further research is necessary to develop automated chart analysis methods that can be effectively applied to these specific areas. In this paper, a comprehensive review of chart analysis methods is presented. The review covers a wide range of chart types, including line charts, bar charts, scatter plots, and includes an in-depth analysis of each method. Additionally, this paper provides a more extensive coverage of chart analysis methods compared to previous studies, making it a valuable resource for researchers and practitioners in the field. Various techniques can be categorized from different aspects, such as chart type, model architecture, learning algorithm, visual feature space, and language modeling. In this paper, different methods are classified from a more technical viewpoint, by considering the approach used for modeling the problem. A taxonomy is proposed which divides the methods into three major categories: rule-based, chart captioning, and chart question-answering approaches. The rule-based approach uses the classical knowledge representation methods for reasoning, which has been diminished by the emergence of deep learning models. Chart capti

The overall level of emissions from the Swiss passenger cars is strongly dependent on the fleet composition. Despite technology improvements, the Swiss passenger cars fleet remains emissions intensive. To analyze the root of this problem and evaluate potential solutions, this paper applies deep learning techniques to evaluate the inter-class (namely micro, small, middle, upper middle, large and luxury class) and intra-class (namely sport utility vehicle and non-sport utility vehicle) differences in carbon dioxide (CO2) emissions. This paper takes full use of novel semi-supervised fuzzy C-means (SSFCM), random forest and AdaBoost models as well as model fusion to successfully classify passenger vehicles and enable segment-based CO2 emission evaluations.

Distributed Energy Resources (DERs) are being integrated into the power market by customers rather than large scale energy suppliers, thereby slowly transforming the centralized, unidirectional market to a decentralized, bidirectional market and transitioning customers into prosumers. Various system architectures are used in the real field to coordinate the energy distribution in the micro/ mini-grids integrated with DERs, all of which have their strengths, weaknesses and challenges. Peer-to-peer (P2P) is an emerging architecture in the field of electrical energy trading and Distributed Generation (DG) management that can be applied in local energy markets. This paper focuses on P2P energy trading, with an in-depth discussion on its various operating algorithms, their principles, characteristics, features and scope through state of art review on P2P. Furthermore, the energy system of Nepal is used as a case study in this paper, and the micro/mini-grids of Nepal and their associated challenges, constraints and opportunities for improvement are discussed. Finally, an energy trading model is proposed to address the problems occurring in the specific case of Nepalese energy market.

The overall level of emissions from the Swiss passenger cars is strongly dependent on the fleet composition. Despite technology improvements, the Swiss passenger cars fleet remains emissions intensive. To analyze the root of this problem and evaluate potential solutions, this paper applies deep learning techniques to evaluate the inter-class (namely micro, small, middle, upper middle, large and luxury class) and intra-class (namely sport utility vehicle and non-sport utility vehicle) differences in carbon dioxide (CO2) emissions. This paper takes full use of novel semi-supervised fuzzy C-means (SSFCM), random forest and AdaBoost models as well as model fusion to successfully classify passenger vehicles and enable segment-based CO2 emission evaluations.

Communication failures and transmission delays are two major issues associated with Wide-Area Damping Controllers (WADCs). While transmission delays have been extensively studied and various solutions have been proposed, little research has been done on communication failures and most of the proposed methods are based on preventive controls. However, in today’s liberalized electricity markets, preventive controls are no longer acceptable and the trend is to use emergency controls instead. This paper proposes a novel emergency control scheme to counteract the loss of remote signals related to the input and to the output of the WADC (i.e. sensor and actuator failures). The proposed scheme is based on a simple criterion, which overcomes the complexity of the previous methods. Modal analysis and time domain simulations are performed to verify the performance of the proposed method. The simulation results show that the proposed method performs well in handling communication failures and can maintain good damping performance. This research work is particularly important in view of the trend towards the wide-scale adoption of wide-area measurement technologies, while the vulnerability to cyber-attacks is increasing.

Machine learning and quantum computing are both areas with considerable progress in recent years. The combination of these disciplines holds great promise for both research and practical applications. Recently there have also been many theoretical contributions of quantum machine learning algorithms with experiments performed on quantum simulators. However, most questions concerning the potential of machine learning on quantum computers are still unanswered such as How well do current quantum machine learning algorithms work in practice? How do they compare with classical approaches? Moreover, most experiments use different datasets and hence it is currently not possible to systematically compare different approaches. In this paper we analyze how quantum machine learning can be used for solving small, yet practical problems. In particular, we perform an experimental analysis of kernel-based quantum support vector machines and quantum neural networks. We evaluate these algorithm on 5 different datasets using different combinations of quantum feature maps. Our experimental results show that quantum support vector machines outperform their classical counterparts on average by 3 to 4% in accuracy both on a quantum simulator as well as on a real quantum computer. Moreover, quantum neural networks executed on a quantum computer further outperform quantum support vector machines on average by up to 5% and classical neural networks by 7%.

One unsolved sub-task of document analysis is mathematical formula detection (MFD). Research by ourselves and others has shown that existing MFD datasets with inline and display formula labels are small and have insufficient labeling quality. There is therefore an urgent need for datasets with better quality labeling for future research in the MFD field, as they have a high impact on the performance of the models trained on them. We present an advanced labeling pipeline and a new dataset called FormulaNet in this paper. At over 45k pages, we believe that FormulaNet is the largest MFD dataset with inline formula labels. Our experiments demonstrate substantially improved labeling quality for inline and display formulae detection over existing datasets. Additionally, we provide a math formula detection baseline for FormulaNet with an mAP of 0.754. Our dataset is intended to help address the MFD task and may enable the development of new applications, such as making mathematical formulae accessible in PDFs for visually impaired screen reader users.

This paper aims to introduce a scientific Semi-Supervised Fuzzy C-Mean (SSFCM) clustering approach for passenger cars classification based on the feature learning technique. The proposed method is able to classify passenger vehicles in the micro, small, middle, upper middle, large and luxury classes. The performance of the algorithm is analyzed and compared with an unsupervised fuzzy C-means (FCM) clustering algorithm and Swiss expert classification dataset. Experiment results demonstrate that the classification of SSFCM algorithm has better correlation with expert classification than traditional unsupervised algorithm. These results exhibit that SSFCM can reduce the sensitivity of FCM to the initial cluster centroids with the help of labeled instances. Furthermore, SSFCM results in improved classification performance by using the resampling technique to deal with the multi-class imbalanced problem and eliminate the irrelevant and redundant features.

One unsolved sub-task of document analysis is mathematical formula detection (MFD). Research by ourselves and others has shown that existing MFD datasets with inline and display formula labels are small and have insufficient labeling quality. There is therefore an urgent need for datasets with better quality labeling for future research in the MFD field, as they have a high impact on the performance of the models trained on them. We present an advanced labeling pipeline and a new dataset called FormulaNet in this paper. At over 45k pages, we believe that FormulaNet is the largest MFD dataset with inline formula labels. Our experiments demonstrate substantially improved labeling quality for inline and display formulae detection over existing datasets. Additionally, we provide a math formula detection baseline for FormulaNet with an mAP of 0.754. Our dataset is intended to help address the MFD task and may enable the development of new applications, such as making mathematical formulae accessible in PDFs for visually impaired screen reader users.

Radial basis function neural networks (RBFs) are prime candidates for pattern classification and regression and have been used extensively in classical machine learning applications. However, RBFs have not been integrated into contemporary deep learning research and computer vision using conventional convolutional neural networks (CNNs) due to their lack of adaptability with modern architectures. In this paper, we adapt RBF networks as a classifier on top of CNNs by modifying the training process and introducing a new activation function to train modern vision architectures end-to-end for image classification. The specific architecture of RBFs enables the learning of a similarity distance metric to compare and find similar and dissimilar images. Furthermore, we demonstrate that using an RBF classifier on top of any CNN architecture provides new human-interpretable insights about the decision-making process of the models. Finally, we successfully apply RBFs to a range of CNN architectures and evaluate the results on benchmark computer vision datasets.

Nepal is a Himalayan country with its 83% of its geography being composed of hills and mountains. Around 22% of the Nepalese population is not receiving electricity through the national power utility and is forced to identify alternative approaches to electrification. The Micro/Mini-Grid (MG) system is one of the promising approaches in terms of cost, reliability and performance for rural electrification, where electrification through national power utility is not techno-economically feasible. However, various issues must be identified and considered during the implementation of MGs in a rural community. In this paper, numerous technical, social and management issues are identified and discussed relating to the implementation and operation of reliable and stable MGs in the Himalayas. To our knowledge, this is the first scientific work that presents a comprehensive review of Himalayan MGs and their associated elements. This article reviews the available research articles, project documents, and Government reports on MG development, from which a clear roadmap is constructed. From the comprehensive study, it is observed that the existing MGs are not adequately designed for the specific area, considering the local resources and local information. Based on the identified issues, some practical and efficient recommendations have been made, so that future MG projects will address the possible problems during the design and implementation phase.

This paper aims to introduce a scientific Semi-Supervised Fuzzy C-Mean (SSFCM) clustering approach for passenger cars classification based on the feature learning technique. The proposed method is able to classify passenger vehicles in the micro, small, middle, upper middle, large and luxury classes. The performance of the algorithm is analyzed and compared with an unsupervised fuzzy C-means (FCM) clustering algorithm and Swiss expert classification dataset. Experiment results demonstrate that the classification of SSFCM algorithm has better correlation with expert classification than traditional unsupervised algorithm. These results exhibit that SSFCM can reduce the sensitivity of FCM to the initial cluster centroids with the help of labeled instances. Furthermore, SSFCM results in improved classification performance by using the resampling technique to deal with the multi-class imbalanced problem and eliminate the irrelevant and redundant features.

Collecting precise real-time information on urban drainage system performance is essential to identify, predict, and manage critical loading situations, such as urban flash floods and sewer overflows. Although emerging low-power wireless communication techniques allow efficient data transfers with great above-ground performance, for underground or indoor applications in a large coverage range are difficult to achieve due to physical and topological limitations, particularly in dense urban areas. In this paper, we first discuss the range limitations of the LoRaWAN standard based on a systematic evaluation of a long-term operation of a sensor network monitoring in-sewer process dynamics. Analyses reveal an-on average-five-fold higher data packet loss for sub-surface nodes, which steadily grows with increasing distance to the gateway. Second, we present a novel LPWAN concept based on the LoRa technology that enhances transmission reliability, efficiency, and flexibility in range-critical situations through meshed multi-hop routing and ensures a precise time-synchronization through optional GPS or DCF77 long-wave time signaling. Third, we illustrate the usefulness of the newly developed concept by evaluating the radio transmission performance for two independent full-scale field tests. Test results show that the synchronous LoRa mesh network approach clearly outperforms the standard LoRaWAN technique with regard to the reliability of packet delivery when transmitting from range-critical locations. Hence, the approach is expected to generally ease data collection from difficult-to-access locations such as underground areas.

Collecting precise real-time information on urban drainage system performance is essential to identify, predict, and manage critical loading situations, such as urban flash floods and sewer overflows. Although emerging low-power wireless communication techniques allow efficient data transfers with great above-ground performance, for underground or indoor applications in a large coverage range are difficult to achieve due to physical and topological limitations, particularly in dense urban areas. In this paper, we first discuss the range limitations of the LoRaWAN standard based on a systematic evaluation of a long-term operation of a sensor network monitoring in-sewer process dynamics. Analyses reveal an-on average-five-fold higher data packet loss for sub-surface nodes, which steadily grows with increasing distance to the gateway. Second, we present a novel LPWAN concept based on the LoRa technology that enhances transmission reliability, efficiency, and flexibility in range-critical situations through meshed multi-hop routing and ensures a precise time-synchronization through optional GPS or DCF77 long-wave time signaling. Third, we illustrate the usefulness of the newly developed concept by evaluating the radio transmission performance for two independent full-scale field tests. Test results show that the synchronous LoRa mesh network approach clearly outperforms the standard LoRaWAN technique with regard to the reliability of packet delivery when transmitting from range-critical locations. Hence, the approach is expected to generally ease data collection from difficult-to-access locations such as underground areas.