Search

With the development of artificial intelligence, we pay widespread attention to the high reliability and interpretability of resulting architectures and algorithms. As one of the main methods of machine learning, the "black-box" mechanism of neural networks lacks interpretability. In order to alleviate the low interpretability, based on belief rule-based inference methodology using the evidential reasoning approach (RIMER), we propose a convolutional rule inference network (CRIN), which is interpretable by involving rules in belief rule base (BRB) under various types of uncertain knowledge representation. Considering the influence of global data distribution on attribute weights, the determination of each attribute may be affected by all other attributes, based on the sigmoid activation function, we determine optimized attribute weights to make sure that each attribute weight is formed by all attribute values. The derivation functions with optimized attribute weights based on evidential reasoning theory also showed to ensure the feasibility of the following learning algorithm. For expressing the influence of local rules input on final output in the network, the network framework and learning algorithm of CRIN based on convolution strategy are proposed. The inference process in RIMER is used as the feedforward mechanism to ensure the interpretability of the network. Meanwhile, the gradient descent method is used as back propagation algorithm to adjust the parameters of rule base to establish a more reasonable BRB. The experimental results and comparative analysis demonstrate that the proposed CRIN has advantages in terms of interpretability and learning capability. This study suggests new trends for interpretable network and uncertain classification issues in the reasoning approach. [ABSTRACT FROM AUTHOR]

The paper presents a concept of an air pollution monitoring system with prediction abilities, based on wireless smart sensors, that takes into account local conditions (microclimate) prevailing in particular areas of the city. In most cases reported in the literature, artificial neural networks (ANNs) are used to predict future pollution levels. In existing solutions of this type, ANNs are trained with generalized datasets common for larger areas, e.g., cities. Our investigations show, however, that conditions may strongly differ even between particular streets in the city, which may impact prediction quality. This results from varying density of urban development, different levels of insolation, airiness, amounts of greenery, etc. As a result, with similar values of ANN input signals, such as current pollution levels, temperature, pressure, etc., the results of the prediction may differ significantly from reality. For this reason, we propose an innovative solution, in which particular sensors are equipped with miniaturized low-power ANNs, trained with datasets gathered directly from their closest environment, without a need for the obtaining of such data from a base station. This may simplify the installation and maintenance process of a network of such sensors. In a further part of this work, we dealt with solutions that enable the reduction of the computational complexity of ANNs in the case of their implementation on specialized integrated circuits. We propose replacing the most complex mathematical operations used in the learning algorithm with simpler solutions. A prototype chip containing the main blocks of such an ANN was also designed. [ABSTRACT FROM AUTHOR]

• Take morphology and shape features into account for granularity representation. • Using interval information granules to achieve system modelling. • A granular model is proposed and used in the field of anomaly detection. • Evaluation indexes are established to quantify the performance. • Greatly reduce the data volume and improve the detection performance. Since time series are characterized by a substantial volume of data, high levels of noise and the correlation between data in the time series attributes, it becomes challenging to mine crucial information from the series and apply it to anomaly detection. In this study, inspired by the concept of information granularity being applied to the process of system modelling, a granular Markov model is proposed for time series anomaly detection. Anomalies are generally caused by the changes in amplitude and shape; in this study we take both the original time series data and their amplitude change data into consideration. First, we utilize an interval information granularity representation based on the principle of justifiable granularity to represent the original time series data in an abstract manner to arrive at the corresponding representation results-- that is, interval information granules. Then, based on the results of the interval information granularity representation and the Fuzzy C-Means (FCM) clustering algorithm, a granular Markov model is developed to produce anomaly scores to quantify possible anomalies. Compared with state-of-the-art methods, experimental studies completed for a large number of datasets demonstrate that the proposed method can significantly improve the anomaly detection process with higher data anomaly resolution. The obtained results are consistent across all datasets. [ABSTRACT FROM AUTHOR]

• Apply dynamic utility calculation framework to mine closed high utility itemsets. • Dramatically reduce the cost of database scans by adopting the P-set structure. • Propose a new CHUIM algorithm, named iEFIM-Closed. • Conduct experiments to measure the performance of the proposed algorithm. The problem of discovering high-utility itemsets (HUIs) in transaction databases, which is an extension of Frequent Itemset Mining, is a commonly encountered mining task. Researchers have proposed algorithms to efficiently mine highly profitable itemsets in customer transaction databases, in which the unit profits of items are fixed. However, this assumption does not reflect the true nature of the utility measure of items in real-life transaction databases, which might vary over time. Moreover, since this important characteristic is ignored by all the current HUI mining algorithms, they are either not applicable to this type of database, or they generate inaccurate results. In addition, the HUI mining algorithms' traditional limitation is that they produce a huge number of HUIs for users. In this paper, we define the problem of mining a lossless, concise and compact representation of HUIs, called closed HUIs (CHUIs), in dynamic unit profit databases. Based on newly defined of utility measure, a novel algorithm, called iEFIM-Closed, is introduced. This relies on this new utility measure, a novel compact database format to reduce the cost of database scans and increase the efficiency of the mining process. Experimental evaluations show that iEFIM-Closed significantly outperforms state-of-the-art algorithms for mining CHUIs on sparse databases with dynamic profit, and has competitive performance in dense databases in terms of runtime, the cost of database scans and scalability. [ABSTRACT FROM AUTHOR]

In remote sensing imagery, segmentation techniques fail to encounter multiple regions of interest due to challenges such as dense features, low illumination, uncertainties, and noise. Consequently, exploiting vast and redundant information makes segmentation a difficult task. Existing multilevel thresholding techniques achieve low segmentation accuracy with high temporal difficulty due to the absence of spatial information. To mitigate this issue, this paper presents a new Rényi's entropy and modified cuckoo search-based robust automatic multi-thresholding algorithm for remote sensing image analysis. In the proposed method, the modified cuckoo search algorithm is combined with Rényi's entropy thresholding criteria to determine optimal thresholds. In the modified cuckoo search algorithm, the Lévy flight step size was modified to improve the convergence rate. An experimental analysis was conducted to validate the proposed method, both qualitatively and quantitatively against existing metaheuristic-based thresholding methods. To do this, the performance of the proposed method was intensively examined on high-dimensional remote sensing imageries. Moreover, numerical parameter analysis is presented to compare the segmented results against the gray-level co-occurrence matrix, Otsu energy curve, minimum cross entropy, and Rényi's entropy-based thresholding. Experiments demonstrated that the proposed approach is effective and successful in attaining accurate segmentation with low time complexity. [ABSTRACT FROM AUTHOR]

Failure mode and effect analysis (FMEA), as a powerful and effective risk assessment tool in reliability and safety analysis, has been extensively applied in different fields to enhance the reliability of a system. However, the traditional FMEA method has exposed some significant defects in practical applications. Furthermore, the risk ranking of failure modes is not convenient for subsequent maintenance strategies, and the interactions between risk criteria are not considered in most multiple criteria decision making (MCDM)-based methods. In this paper, we construct a novel FMEA model based on criteria importance through the inter-criteria correlation (CRITIC) method and three-way decisions to improve the performance of the conventional FMEA method and aid making a maintenance plan. Firstly, we introduce the decision-theoretic rough sets (DTRSs) into an interval 2-tuple linguistic (I2TL) environment to define interval 2-tuple linguistic decision-theoretic rough sets (I2TLDTRSs). Secondly, the CRITIC method is employed to calculate the weight of risk criteria. Thirdly, the conditional probability is obtained by a similarity-based method with the help of the ideas of the TOPSIS approach, and a relative loss function is designed by introducing risk avoidance coefficient. Fourthly, the classification of failure modes is derived according to the three-way decision rules. Finally, two numerical examples of the train door system are given to verify the effectiveness and practicality of the presented model. The presented model both extends the theory and application of three-way decisions and provides a solution beneficial for making a maintenance strategy to reduce the risk of failure modes. [ABSTRACT FROM AUTHOR]

Contrast sets are used in many knowledge-based systems to capture data patterns relevant to a target variable. While they have many advantages such as being highly interpretable, they do not come with a similarity measure or feature vectors for downstream tasks such as regression or classification. To address these disadvantages, we propose Con2Vec (Con trast set to Vec tor), a method to embed contrast sets into a low-dimensional continuous vector space. Con2Vec defines two novel similarity and co-occurrence contexts for a contrast set, and then leverages a neural embedding model to learn low-dimensional continuous vectors (aka embeddings) for contrast sets. We further apply contrast set embeddings to construct the feature vectors for transactional data. We extensively evaluate our method Con2Vec on four real-world datasets, compared against state-of-the-art embedding and non-embedding methods where the results demonstrate the clear advantages of our method. [ABSTRACT FROM AUTHOR]

Due to the high data volume and non-stationarity of time series data, it is very difficult to directly use the original data for anomaly detection. In this study, a novel framework of anomaly detection is proposed, whose intent is to capture more detailed data of time series' shape and morphology characteristics by data representation to carry out anomaly detection. First, high-order differences and intervals are employed to realize data representation, and then such rectangles and cubes are constructed with the results of data representation for similarity measurement and anomaly detection. Compared with existing state-of-the-art methods, based on the experimental studies completed on large amount of datasets, the methods proposed in this framework are effective in detecting anomalies caused by changes in shape and amplitude. Meanwhile, it can detect anomalies with higher accuracy and better performance of data anomaly resolution. [ABSTRACT FROM AUTHOR]

In this paper, we study on how to achieve balance between minimizing the objectives, satisfying the constraints, and avoiding the population to stuck at locally optimal or locally feasible regions. We propose a new constraints handling technique based on diversity distance measure for constrained multiobjective optimization problems. We innovatively measure all effects of the constraints on the objectives by the interaction effects and direct effects of constraints on the objectives. The interactive effects of constraints on the objective functions are firstly expressed and quantified by using Mahalanobis distance and computing the value of the distance measure by involving a Lebesgue measure. Based on interactive and direct effects, we propose a new modified mechanism of objective space, called objective space modified mechanism based on diversity distance measure, so that the good infeasible solution could be more effectively used to find the optimal solution. The Pull and Push search is used to adaptively adjust the location of the Pareto front, which could preclude the population from being stuck at some locally optimal or locally feasible regions and could decrease time complexity. Comprehensive experiments completed for several benchmark problems demonstrate the competitiveness of the proposed algorithm, in comparison to the existing state-of-art constrained evolutionary multiobjective optimization. [ABSTRACT FROM AUTHOR]

The mobile wireless networks (MWNs) are flexible and scalable. They have been widely used in large-scale monitoring, search and rescue tasks, pursuit and escape operations and other scenarios. However, due to the high cost of wireless sensors in MWNs, the cost of increasing wireless sensor nodes is relatively high, so it is generally monitored by the limited nodes moving in the monitoring range. Having this in mind, we design an original algorithm to improve the monitoring efficiency of mobile wireless nodes, named Distributed Guidance Anti-flocking Algorithm (DGAA). Because most of the dynamic coverage algorithms are difficult to adapt to the complex operating environment. Furthermore the high overlapping rate of the wireless nodes is easy to cause resource waste. In light of these factors, the DGAA has a low overlapping rate of mobile wireless nodes in the complex environment with obstacles. DGAA mimics the predatory behavior of solitary animals. These animals can go out on their own to catch prey, and they do not leave their homes. At the same time, they can prevent the invasion of foreign individuals and ensure their maximum profits. The experiment is simulated on a map of 50m*50m, and the results are compared with those produced by competitive algorithms. Experimental results show that DGAA exhibits better performance than other algorithms. [ABSTRACT FROM AUTHOR]

• We study a sequential defender-attacker game; • Multiple strategies and targets are considered; • Learning and counter-learning behaviors are modeled; • The interaction between targets resulting in additional utility is allowed; • Comparative studies are implemented to see how parameters affect the equilibrium. Appropriate resource allocation among multiple targets plays a vital role in the defender-attacker game. Traditional methods seldom consider the optimal resource allocation with more than one strategy among multiple targets. Accordingly, we assume that the defender moves first, allocating resources in deploying different false targets and strengthening the genuine ones among multiple targets. The attacker moves after observing the defender's move, distributing resources in identifying false targets and then attacking the genuine ones concerning multiple targets. Each player aims to maximize their own overall expected utility given any of the other player's move. The optimal resource distribution between target intelligence/protection for the defender and target identification/attack efforts among multiple targets for the attacker is studied. We model and obtain analytical equilibrium results under various parameter settings. Our results show that increased intelligence (impact) contest intensities lead to lower (higher) attacker resources in intelligence (attack). Moreover, the zero-sum game where both players share the same preferences without additional payoff consideration results in the worst losses for the defender. The forgoing method offers an original way to approach the resource allocation problems in security defense fields. [ABSTRACT FROM AUTHOR]

Studying the well-known phenomenon "aurora" plays a pivotal role in investigating the solar–terrestrial coupling mechanism. A special auroral spectrograph in Antarctic Zhongshan Station constitutes a auroral observation joint system with satellite-borne sensors of the Defense Meteorological Satellite Program. Multipoint observation by this system provides more essential information for relevant studies than single observation by each instrument, but also results in a multifold increased volume of data that are difficult to be either stored or transmitted. To address this difficulty, we develop a clustering-based, generic lossless data compression framework that combines the usage of various ultimate compressors with a hierarchical clustering algorithm to exert the strength of all the compressors in data reduction. This framework achieves an always-best compression performance for different-sized datasets with a reasonable time consumption, which promises the design of pipelines using it for real-time data transmission. [ABSTRACT FROM AUTHOR]

Burn-through point (BTP) is an essential parameter in the iron ore sintering process. Operators usually judge whether the current production is stable by monitoring the BTP. It comes with significant application prospects to predict the BTP accurately. A prediction model of the BTP with fuzzy time series is designed in this paper. First, the fuzzy time series prediction method with the Fuzzy C-Means clustering is presented as the core modeling method. A prediction model of the response is constructed to obtain a timely response to the current BTP. The prediction model of the difference is established to estimate the present unmeasurable disturbance on the BTP. Then, a hybrid prediction model is built, which realizes the composition of these two models by an adjustment factor. Finally, a series of experiments is carried out using the raw time series data from an iron and steel plant. The experimental result shows that the designed model has better prediction performance for the BTP than existing models, which is an advantage resulting from the hybrid structure and the fuzzy time series prediction model with the Fuzzy C-Means clustering. This prediction model of the BTP implies the foundation for the stable control of the iron ore sintering process. • A fuzzy time series prediction model with the Fuzzy C-Means clustering is designed. • A hybrid prediction model of the burn-through point is developed. • The proposed model has a higher prediction accuracy than the existed models. [ABSTRACT FROM AUTHOR]