Search

A taxonomy that classifies ensemble models in the literature is presented.Surface and deep features integration is explored to improve classification.Several ensembles of classifiers and features are proposed and evaluated.Performance of the proposed models is evaluated on several sentiment datasets. Deep learning techniques for Sentiment Analysis have become very popular. They provide automatic feature extraction and both richer representation capabilities and better performance than traditional feature based techniques (i.e., surface methods). Traditional surface approaches are based on complex manually extracted features, and this extraction process is a fundamental question in feature driven methods. These long-established approaches can yield strong baselines, and their predictive capabilities can be used in conjunction with the arising deep learning methods. In this paper we seek to improve the performance of deep learning techniques integrating them with traditional surface approaches based on manually extracted features. The contributions of this paper are sixfold. First, we develop a deep learning based sentiment classifier using a word embeddings model and a linear machine learning algorithm. This classifier serves as a baseline to compare to subsequent results. Second, we propose two ensemble techniques which aggregate our baseline classifier with other surface classifiers widely used in Sentiment Analysis. Third, we also propose two models for combining both surface and deep features to merge information from several sources. Fourth, we introduce a taxonomy for classifying the different models found in the literature, as well as the ones we propose. Fifth, we conduct several experiments to compare the performance of these models with the deep learning baseline. For this, we use seven public datasets that were extracted from the microblogging and movie reviews domain. Finally, as a result, a statistical study confirms that the performance of these proposed models surpasses that of our original baseline on F1-Score.

Present and formally define the Temporal Multi-Agent Resource Allocation diagnosis problem.Propose an efficient solution to this problem by a novel encoding to Boolean Satisfiability (SAT).Demonstrate empirically the scalability of the proposed approach. In the not-so-far future, autonomous vehicles will be ubiquitous and, consequently, need to be coordinated to avoid traffic jams and car accidents. A failure in one or more autonomous vehicles may break this coordination, resulting in reduced efficiency (due to traffic load) or even bodily harm (due to accidents). The challenge we address in this paper is to identify the root cause of such failures. Identifying the faulty vehicles in such cases is crucial in order to know which vehicles to repair to avoid future failures as well as for determining accountability (e.g., for legal purposes). More generally, this paper discusses multi-agent systems (MAS) in which the agents use a shared pool of resources and they coordinate to avoid resource contention by agreeing on a temporal resource allocation. The problem we address, called the Temporal Multi-Agent Resource Allocation (TMARA) diagnosis problem (TMARA-Diag), is to find the root cause of failures in such MAS that are caused by malfunctioning agents that use resources not allocated to them. As in the autonomous vehicles example, such failures may cause the MAS to perform suboptimally or even fail, potentially causing a chain reaction of failures, and we aim to identify the root cause of such failures, i.e., which agents did not follow the planned resource allocation. We show how to formalize TMARA-Diag as a model-based diagnosis problem and how to compile it to a set of logical constraints that can be compiled to Boolean satisfiability (SAT) and solved efficiently with modern SAT solvers. Importantly, the proposed solution does not require the agents to share their actual plans, only the agreed upon temporal resource allocation and the resources used at the time of failure. Such solutions are key in the development and success of intelligent, large, and security-aware MAS.

An anomaly detection algorithm to identify AIS on-off switching is proposed.The algorithm exploits the AIS message Received Signal Strength Indicator.Machine Learning algorithms are used to build normality models.AIS reception is characterized by using real word data.The methodology is scalable from one station to a network of receivers. The Automatic Identification System (AIS) is a ship reporting system based on messages broadcast by vessels carrying an AIS transponder. The recent increase of terrestrial networks and satellite constellations of receivers is making AIS one of the main sources of information for Maritime Situational Awareness activities. Nevertheless, AIS is subject to reliability and manipulation issues; indeed, the received reports can be unintentionally incorrect, jammed or deliberately spoofed. Moreover, the system can be switched off to cover illicit operations, causing the interruption of AIS reception. This paper addresses the problem of detecting whether a shortage of AIS messages represents an alerting situation or not, by exploiting the Received Signal Strength Indicator available at the AIS Base Stations (BS). In designing such an anomaly detector, the electromagnetic propagation conditions that characterize the channel between ship AIS transponders and BS have to be taken into consideration. The first part of this work is thus focused on the experimental investigation and characterisation of coverage patterns extracted from the real historical AIS data. In addition, the paper proposes an anomaly detection algorithm to identify intentional AIS on-off switching. The presented methodology is then illustrated and assessed on a real-world dataset.

A new initialization technique is proposed to improve the performance of K-means.A data transformation approach is proposed to solve empty cluster problem.An efficient method is proposed to estimate the optimal number of clusters.Proposed clustering method provides more accurate clustering results. Clustering provides a knowledge acquisition method for intelligent systems. This paper proposes a novel data-clustering algorithm, by combining a new initialization technique, K-means algorithm and a new gradual data transformation approach to provide more accurate clustering results than the K-means algorithm and its variants by increasing the clusters coherence. The proposed data transformation approach solves the problem of generating empty clusters, which frequently occurs for other clustering algorithms. An efficient method based on the principal component transformation and a modified silhouette algorithm is also proposed in this paper to determine the number of clusters. Several different data sets are used to evaluate the efficacy of the proposed method to deal with the empty cluster generation problem and its accuracy and computational performance in comparison with other K-means based initialization techniques and clustering methods. The developed estimation method for determining the number of clusters is also evaluated and compared with other estimation algorithms. Significances of the proposed method include addressing the limitations of the K-means based clustering and improving the accuracy of clustering as an important method in the field of data mining and expert systems. Application of the proposed method for the knowledge acquisition in time series data such as wind, solar, electric load and stock market provides a pre-processing tool to select the most appropriate data to feed in neural networks or other estimators in use for forecasting such time series. In addition, utilization of the knowledge discovered by the proposed K-means clustering to develop rule based expert systems is one of the main impacts of the proposed method.

The identification and prediction of situations in smart houses.An approach developed with the principles of reactive and proactive behavior.A case study presented with sensors and appliances.The proactivity and reasoning about uncertainty supported by probabilistic ontology. Considering that the world population is aging, health-support issues are in evidence, and many dangerous situations concerning users in their living environment may arise. This paper presents an approach to allow Ambient Assisted Living (AAL) systems to identify and predict situations that may endanger users in their living environment. Elderly people generally have difficulties to perform daily activities as using some electronic device, cooking, dressing among others. This paper offers an alternative to acquire comprehension of a former person's behavioral providing this knowledge when cognitive impairments will occur. This is a significant improvement to traditional expert systems applied to AAL as the handling of the knowledge base is adapted to the failing users behavior. We consider the user profile dynamics where new and different situations not initially planned in the development of the systems may occur. This approach was developed considering the principles of reactive and proactive behavior and the extensibility capacity of the system. Systems developed following the above-described principles can react to the current situations and anticipate proactively a dangerous situation eliminating it or controlling its impact. The approach is evaluated and verified by a case study where unwanted, dangerous situations were created and subsequently detected. The case study is based on a simulated smart house fulfilled with sensors and appliances. An application was build based on the developed approach, and it was capable of choosing the most appropriate actions to solve the unwanted situations. The main contributions of the present work are the identification and prediction of situations in smart houses, an approach developed with the principles of reactive and proactive behavior validated with a case study and supported by a probabilistic ontology.

With the rapid development of location-based services, a large amount of moving object trajectory data has been generated. Mining users’ stay area from moving trajectory data can not only improve location-based service, but also take precautions to ensure personal safety to better promote the development of location-based services. However, the research on trajectory of indoor moving objects is still in its infancy, and there is a lack of relevant research in this field. It is urgent to meet the needs of the location service industry. We have done some novel work on the stay area in indoor trajectories. Firstly, a novel algorithm named SVBDSA is proposed for the stability value-based discovery of stay area. The stability value is calculated from the distance and velocity between the two points. By comparing the stability value with the threshold value, the two points are judged whether they are in a stable state, and then the point in the stable state is extended. Thus, the stay area in the moving trajectories can be discovered. Secondly, this paper proposes an improved algorithm named T - O P T I C S , which adds the time attribute and puts forward the concept of difference value, so that abnormal stay points can be detected in the discovered stay area. Finally, the trajectory generation tool Vita is used to conduct experiments and analysis and the effectiveness of the proposed algorithms in this paper is verified. Discovery of stay area is not only a new and difficult scientific issue that has not been studied in the field of indoor trajectories of the moving object, but also has important academic value and research significance in location-based service industry.

The problem of influence maximization in social networks has been widely investigated, but most previous studies have usually ignored the dynamic nature of propagation and the effects of local aggregation factors on diffusion. This paper presents a Dynamic algorithm based on cohesive Entropy for Influence Maximization (DEIM), the goal of which is to find the most influential nodes in social networks. Firstly, the Community Overlap Propagation Algorithm based on Cohesive Entropy (CECOPA) is put forward for the discovery of overlapping communities in networks, and potential nodes in the gathering area are selected to construct the candidate seed set. Then, the Optional Dynamic influence Propagation algorithm (ODP) is designed based on narrowing down the selection range of seeds. It utilizes a variety of entropy calculations to obtain the cohesive power between neighboring nodes and then determines whether the node has the ability to become a propagable pioneer of another node; thus, information continues to diffuse effectively. Finally, via many times experiments on several data sets, it is confirmed that the proposed DEIM algorithm in this paper can successfully affect the ideal number of users in different scenarios.

The paper considers a problem of forecasting of news feeds content. Analysis of existing approaches to this problem solution reveals the need for development of methods with enhanced forecasting capabilities. A method is proposed with expanded accounting for space and time relationships of the processed data. The method is revealed through an example of a neural network forecasting system that implements it. Implementation includes data retrieval from news feeds, their special preprocessing, coding and forecasting of words sets and their interconnections, followed by highlighting news topics and describing the of news feeds content. Some variants of stream recurrent neural networks with spiral layer structures were investigated with due regard to their forecasting capabilities under direction and strength control of the associative call of signals from the network memory. The paper also presents and discusses experimental results, a description of the methodological contribution and recommendations on the method practical application.

This paper provides a multi-stage medical scheme selection process to obtain the suitable scheme in the referral system. Firstly, considering the uncertainty of the experts and physicians, the preference relations (PRs) under the probabilistic linguistic term environment are introduced to the system. Then, this paper checks the consistency of the PRs and selects out the inconsistent PRs. For the inconsistent one, this paper provides alternative evaluations by repairing the most inconsistent element. Then, let experts choose from the alternative evaluation set. Next, by introducing the Bonferroni mean (BM) operator and Choquet integral, this paper integrates the historical data and the evaluations with stage weights to obtain the comprehensive assessment. The weights are calculated by the provided discrete and continuous stage weight functions or the programming model. Finally, the medical scheme selection process for the lung cancer and its simulation experiment are investigated to demonstrate the effectiveness and application of provided methods. The simulation experiments on sensitivity analysis and comparative analysis with existing methods are also conducted to validate the provided method.

This paper presents a small-sample faulty line detection method for a small current grounded system (SCGS) using improved generative adversarial networks (GANs) to overcome the problems of low accuracy and poor economy associated with traditional methods. First, we collected real-time fault data of the dispatching system as the original sample and used the linear interpolation and normalization methods to preprocess the data. We extracted and labeled the features of the original sample to construct a feature vector that represents the single-phase grounded fault. Then, to solve the problem of insufficient faulty samples, this paper used the GAN to generate synthetic samples to expand the training sample size. To stabilize the GAN training, the Wasserstein distance was introduced to improve the GAN loss function (WGAN). Finally, for the synthetic samples, we used Batch Normalization (BN) and an Early Stopping algorithm to train the lightweight depthwise separable convolutional network (DSCN) classifier and obtain a faulty line detection model with fast convergence and high accuracy. With the data of an actual dispatching system in China as an example, the proposed method determined the faulty line within 100–200 ms with an accuracy of more than 96%, which is approximately 26% higher than the accuracy of traditional methods. Compared with mainstream machine learning algorithms, the proposed method realizes the self-extraction of faulty features and simplifies the algorithm. Moreover, the accuracy of the proposed method is much higher than that of other machine learning algorithms on small samples. It is worth noting that this method needs only the existing real-time dispatching data and does not need to install additional data acquisition and fault detection equipment, eliminating the complicated equipment installation and debugging process and greatly reducing equipment costs human operations and maintenance costs. The proposed method has been implemented in an actual power system in China for two years. The results show that it is highly accurate and cost-effective, performs well in real time, and has bright prospects for future applications.

Signed social networks which are using both negative and positive links are becoming a popular form of social networks. This paper seeks to gain insight into how a user creates a positive or negative connection towards other users. The sign of connections between the users can be predicted efficiently and reliably using a newly proposed metric. In this paper, a new representation which is named Inverse square Metric is proposed. Inverse square metrics is inspired from Inverse square law which uses node properties and distance-based metrics to represent a connection in social networks. Inverse square metric measures the importance and intensity of a pair of nodes using node properties and defines the distance between two nodes as penalty. 16 variant of networks are created to compute the distance between two nodes in the network. In this study, two main contributions are made. First, a new metric to represent edges is proposed which uses both neighbourhood and distance-based link prediction measures. In addition, a unified framework for sign and link prediction problem based on the new representation is proposed. The experimental results on a group of six large networks including Amazon, Facebook, arXiv ASTRO-PH, Epinions, Slashdot, and Wikipedia prove the effectiveness of the proposed method.

Peer review is an essential part of scientific communications to ensure the quality of publications and a healthy scientific evaluation process. Assigning appropriate reviewers poses a great challenge for program chairs and journal editors for many reasons, including relevance, fair judgment, no conflict of interest, and qualified reviewers in terms of scientific impact. With a steady increase in the number of research domains, scholarly venues, researchers, and papers in academia, manually selecting and accessing adequate reviewers is becoming a tedious and time-consuming task. Traditional approaches for reviewer selection mainly focus on the matching of research relevance by keywords or disciplines. However, in real-world systems, various factors are often needed to be considered. Therefore, we propose a multilayered approach integrating Topic Network, Citation Network, and Reviewer Network into a reviewer Recommender System (TCRRec). We explore various aspects, including relevance between reviewer candidates and submission, authority, expertise, diversity, and conflict of interest and integrate them into the proposed framework TCRRec. The paper also addresses cold start issues for researchers having unique areas of interest or for isolated researchers. Experiments based on the NIPS and AMiner dataset demonstrate that the proposed TCRRec outperforms state-of-the-art recommendation techniques in terms of standard metrics of precision@k, MRR, nDCG@k, authority, expertise, diversity, and coverage.

With image object detection techniques progressing, video object detection attracts a lot of attention more than ever before. However, the performance of detecting the object from a video suffers a lot as object occlusion can always lead to the object in the presence of appearance deterioration. To deal with this problem, the temporal correlation of video sequences is often used to reduce the effects of object occlusion in video object detection. In this paper, an optical-flow-feature fusion-based video object detection method is proposed with consideration of temporal coherence among video frames. To further reduce computational complexity, this paper also proposes a packet video processing method. Specifically, video frames are grouped first, and all frames in current group share the same optical flow feature map by feature fusion. Then the Target Image can be formed with object information enhanced by fusing the shared feature map with current frame. The proposed method gives rise to effective background information masking, so that the object detection network can focus more on the foreground object. This method effectively improves the object detection performance and the scene migration performance. Experimental results prove that the proposed method significantly improves the detection accuracy to 78.6% on ImageNet VID. In addition, VGG and ResNet are compared to further verify the effectiveness of the proposed method, and the results can be a persuasive evidence with the highest detection accuracy and acceptable time consumption.

In recent years, the rapid development of mobile network and wireless sensor technology has brought opportunities to change the way of the existing taxi business operation. How to improve the operation revenues of taxi drivers has become a topic worthy of research. This paper analyzes and mines taxi operation data to provide taxi drivers with personalized sequence recommendation services, thereby increasing their expected revenues. Different from previous works, the proposed method in this paper recommends a series of future operation orders for taxi drivers, instead of recommending several discrete locations for the current order. In this paper, firstly, by performing spatial-temporal clustering on the origins and destinations of passengers, the spatial and temporal distribution characteristics of passengers in the city are identified. Secondly, the origin of the current passenger is used as the root node to construct a top-N high-utility sequence tree, and this process can be divided into two processes: top-down building tree and bottom-up sorting path utility. The two pruning strategies of node utility and path utility are used to reduce the generation of candidate sets. Finally, a series of potential orders based on dynamic context are recommended to taxi drivers, so as to maximize the expected revenues of taxi drivers. The experimental results demonstrate that there is a close relationship between taxi drivers’ operation behavior patterns and their revenues. The proposed system framework and algorithm in this paper can effectively mine global and long-term top-N high-utility operation patterns.

This paper describes a new hybrid approach, based on modular artificial neural networks with fuzzy logic integration, for the diagnosis of pulmonary diseases such as pneumonia and lung nodules. In particular, the proposed approach analyzes medical images, which are digitized chest X-rays, focusing on a classification method based on descriptors, such as grayscale histogram features, gray-level co-occurrence matrix (GLCM) texture-based features, and local binary pattern texture features. Then, to perform feature reduction, a multi-objective genetic algorithm is used to obtain an optimized neuro-fuzzy classifier, which is able to classify the pathology found in the analyzed chest X-ray. The main contribution of this paper is the proposed modular neural network approach, which divides features to achieve specialized analysis in the modules for digital image analysis and classification. The proposed approach achieves high classification accuracy after evaluating the neuro-fuzzy model with three large datasets of chest X-rays.

Stock time series forecasting is a universal purpose of academic researchers, even a slight improvement in the accuracy of the forecast may have a fabulous impact on participants’ trading decisions. An appropriate forecasting method will help investors to obtain a reasonable portfolio, to some extent avoid potential financial risks, which makes our research more profound. To provide stock market participants and researchers with better prediction results, in this paper, a modified modeling procedure is proposed which aims to heighten the function of the k-nearest neighbors (KNN) method in financial time series prediction. The key idea is to combine Ensemble empirical mode decomposition (EEMD) method with Multidimensional k-nearest neighbors-time series prediction with invariance (MKNN–TSPI). This new time series forecasting framework is called EEMD–MKNN–TSPI. Take note of the facilitation of the experiment, we present an example of two-dimensional stock series, the modified method is utilized to forecast the opening and closing prices of NAS, DJI, S&P 500, Russell 2000 and other stocks simultaneously. Besides, we compare our method with EEMD–MKNN and MKNN–TSPI, the experimental results indicate that the proposed model outperforms EEMD–MKNN model and MKNN–TSPI model. At the end of the paper, we select some stock sequences and traffic sequences to conduct a series of experiments, the purpose is to explore the impact of data types and complexity on the prediction results about the EEMD–MKNN–TSPI method.

This paper provides a comprehensive survey of various techniques used in computer vision for the automatic identification of plants with the help of leaf images. The extracted information is used by the botanists to identify different species of plants and use their medicinal or other properties. With the upsurge of human interference, the number of plants seems to decrease, but their automatic identification can lead to conservation. The Leaf images may be acquired by a phone camera or a digital camera mounted on a tripod stand. The leaves may be covered by dirt, shadows, or hidden under other leaves. Real-life applications based on the automatic identification of plants can successfully identify even similar-looking plant leaves in all environmental conditions. This paper provides a state-of-the-art review of different leaf extraction techniques which are categorized according to the features of leaf used and their pros and cons. We also discuss and compare the various classifiers used in the identification process. The conclusion of the paper also provides different areas of improvement and future work.

Knowledge-based systems reason over some knowledge base. Hence, an important issue for such systems is how to acquire the knowledge needed for their inference. This paper assesses active learning methods for acquiring knowledge for "static code warnings". Static code analysis is a widely-used method for detecting bugs and security vulnerabilities in software systems. As software becomes more complex, analysis tools also report lists of increasingly complex warnings that developers need to address on a daily basis. Such static code analysis tools are usually over-cautious; i.e. they often offer many warnings about spurious issues. Previous research work shows that about 35% to 91% of warnings reported as bugs by SA tools are actually unactionable (i.e., warnings that would not be acted on by developers because they are falsely suggested as bugs). Experienced developers know which errors are important and which can be safely ignored. How can we capture that experience? This paper reports on an incremental AI tool that watches humans reading false alarm reports. Using an incremental support vector machine mechanism, this AI tool can quickly learn to distinguish spurious false alarms from more serious matters that deserve further attention. In this work, nine open-source projects are employed to evaluate our proposed model on the features extracted by previous researchers and identify the actionable warnings in a priority order given by our algorithm. We observe that our model can identify over 90% of actionable warnings when our methods tell humans to ignore 70 to 80% of the warnings., Comment: Accepted to Expert Systems with Applications

The current paper investigates the problem of multimodal named entity recognition from Twitter data. Named entity recognition (NER) is an important task in natural language processing and has been carefully studied in recent decades. NER from tweets is particularly challenging because of 1) tweets are limited in length, 2) contains noisy text, and 3) contains hashtags. Moreover often tweets are associated with images and hyperlinks. Existing works on tweet-NER mostly concentrate on multimodal deep learning based models neglecting the use of hand-crafted features and usage of hyperlinks. The current paper investigates the incorporation of hand-crafted features extracted from different modalities like images, hyperlinks while extracting named entities from tweet-text. A large set of hand-crafted features are extracted from different modalities (images, hyperlinks) and those are added with the features extracted by a hybrid deep-neural model, bi-directional LSTM and CNN, followed by a conditional random field to perform this task. Several variants of these models in association with different hand-crafted feature sets are designed. Extensive experimentations on a multimodal Twitter data (containing text, images and urls) illustrate that character level hand-crafted features significantly improve the performance of the systems. In a part of the paper, results of the proposed models are also shown on a standard NER dataset, CoNLL 2003 dataset.

In some applications, a relation between a response variable and one or more explanatory variables (referred as a “profile”) characterizes the quality of a process. Profile monitoring is commonly performed through statistical methods, while machine learning schemes have not received much attention in this regard. In this paper, a control chart based on Artificial Neural Networks (ANN) is proposed to monitor linear profiles in phase II. In the proposed control chart, some novel run rules as the major contribution of this paper are also used to enhance the efficiency of the control chart and for faster detection of shifts. Simulation results revealed a good performance of the proposed control chart based on average run length (ARL) criterion. Further, a systematic ANN-based diagnostic procedure was proposed to identify which parameter has changed in the process. Finally, the implementation of the proposed scheme was illustrated through a real calibration example from the field of chemical engineering.

Exploration, the process of visiting a new region in a search space, and exploitation, the process of searching in the neighborhood of previously visited regions, are two centerpieces of any metaheuristic algorithm. It is a common belief that good results can be obtained only if there is a good balance between exploration and exploitation. Hence, there is an urgent need to control the balance between exploration and exploitation in a direct manner. But, currently, direct measures of exploration and exploitation are almost non-existent, and researchers rely on indirect measures of exploration and exploitation, such as diversity, entropy, and fitness improvements. To remedy this situation, in this paper, a novel direct measure of exploration and exploitation is proposed that is based on attraction basins — parts of a search space where each part has its own point called an attractor, to which neighboring points tend to evolve. Each search point can be associated with a particular attraction basin. If a newly generated search point belongs to the same attraction basin as its parent then the search process is identified as exploitation, otherwise as exploration. In this paper, a new technique to compute attraction basins is presented, as well as a novel direct measure ( E x p B a s ) of exploration and exploitation based on attraction basins. On the selected set of unimodal and multimodal optimization problems it is shown that the newly proposed direct measure of exploration and exploitation is more accurate than our previously proposed direct measure ( E x p D i s t ), as well as the common indirect measure based on diversity (Diversity).

With the development of mobile Internet, microblog has become one of the most popular social platforms. The enormous user-generated microblogs have caused the problem of information overload, which makes users difficult to find the microblogs they actually need. Hence, how to provide users with accurate microblogs has become a hot and urgent issue. In this paper, we propose an approach of hybrid microblog recommendation, which is developed on a framework of deep neural network with a group of heterogeneous features as its input. Specifically, two new recommendation strategies are first constructed in terms of the extended user-interest tags and user interest topics, respectively. These two strategies additionally with the collaborative filtering are employed together to obtain the candidate microblogs for final recommendation. Then, we propose the heterogeneous features related to personal interests of users, interest in authors and microblog quality to describe the candidate microblogs. Finally, a deep neural network with multiple hidden layers is designed to predict and rank the microblogs. Extensive experiments conducted on the datasets of Sina Weibo and Twitter indicate that our proposed approach significantly outperforms the state-of-the-art methods. The code and the two datasets of this paper are publicly available at GitHub.

This paper presents an algorithmic method that hybridizes solution procedures with an optimization model for solving the problem of scheduling and allocating changeover tasks. This problem involves a limited number of changeover workers with different skills and a set of sequence-dependent changeover tasks on a multi-machines production line, where executing each task requires a particular type and number of workers. For solving this allocation and scheduling problem, the paper identifies and investigates three priority rules expressed as objective functions to minimize the total changeover time and maximize the workers' utilization while satisfying changeover task-sequence-dependency and worker-limit constraints. For illustrating the applicability and suitability of the proposed approach, a real-world case problem of a 10-machine production line is considered for scheduling and allocating 34 sequence-dependent changeover tasks to 4 workers. The computational effort and time for optimizing each of the three identified objective functions using the proposed method were significantly undersized. The key results are the effective changeover time and workers' utilization provided by the proposed hybrid approach.

Priority rules are applied in many commercial software tools for scheduling projects under limited resources because of their known advantages such as the ease of implementation, their intuitive working, and their fast speed. Moreover, while numerous research papers present comparison studies between different priority rules, managers often do not know which rules should be used for their specific project, and therefore have no other choice than selecting a priority rule at random and hope for the best. This paper introduces a decision tree approach to classify and detect the best performing priority rule for the resource-constrained project scheduling problem (RCPSP). The research relies on two classification models to map project indicators onto the performance of the priority rule. Using such models, the performance of each priority rule can be predicted, and these predictions are then used to automatically select the best performing priority rule for a specific project with known network and resource indicator values. A set of computational experiment is set up to evaluate the performance of the newly proposed classification models using the most well-known priority rules from the literature. The experiments compare the performance of multi-label classification models with multi-class classification models, and show that these models can outperform the average performance of using any single priority rule. It will be argued that this approach can be easily extended to any extension of the RCPSP without changing the methodology used in this study. info:eu-repo/semantics/publishedVersion

This paper deals with the role of alternative optimal solutions existing in the data envelopment analysis (DEA) models for cross-efficiency evaluation in portfolio selection. The paper shows that incorporating alternative optimal solutions for constructing cross-efficiency matrix improves the result of the mean-variance portfolio selection method. This improvement means that building portfolios with lower risk and higher expected return is possible when alternative optimal solutions are considered. The proposed method in this paper is applied to stock portfolio selection in the Tehran stock market.

This paper presents a systematic procedure to fairly compare experimental performance scores for machine learning methods for fault diagnosis based on vibration signals. In the vast majority of related scientific publications, the estimated accuracy and similar performance criteria are the sole quality parameter presented. However, the experimental design giving rise to these results is mostly biased, based on unacceptably simple validation methods and on recycling identical patterns in test data sets, previously used for training. Moreover, the methods in general overfit their hyperparameters, introducing additional overoptimistic results. In order to remedy this defect, we critically analyse the usual training-validation-test division and propose an algorithmic guideline in the form of a validation framework. This allows a well defined comparison of experimental results. In order to illustrate the ideas of the paper, the Case Western Reserve University Bearing Data benchmark is used as a case study. Four distinct classifiers are experimentally compared, under gradually more difficult generalization tasks using the proposed evaluation framework: K-Nearest-Neighbor, Support Vector Machine, Random Forest and One-Dimensional Convolutional Neural Network. An extensive literature review suggests that most vibration based research papers, particularly for the Case Western Reserve University Bearing Data, use similar patterns for training and testing, making their classification an easy task.

This paper investigates the problem scheduling a set of jobs on parallel batch processing machines with different capacities and non-identical processing powers for minimizing the makespan and the total energy consumption, where the jobs have non-identical sizes, dynamical arrival time and different processing time. To address the bi-objective optimization problem, a three-populations co-evolutionary algorithm is proposed, which is based on exploration and coordination searches among three colonies. For guaranteeing the diversity of solutions, an adaptive search strategy based on the largest angle among adjacent solutions is designed, and a new method is proposed to select ants to update pheromone trails for improving the convergence of solutions. Finally, the proposed algorithm is compared with the existing multi-objective algorithms through extensive simulated experiments, and the simulated results are statistically analyzed. And the experimental results show that the proposed algorithm outperforms all the compared algorithms, which verify the validity of the algorithm proposed in this paper.

Integrating return prediction of traditional time series models in portfolio formation can improve the performance of original portfolio optimization model. Since machine learning and deep learning models have shown overwhelming superiority than time series models, this paper combines return prediction in portfolio formation with two machine learning models, i.e., random forest (RF) and support vector regression (SVR), and three deep learning models, i.e., LSTM neural network, deep multilayer perceptron (DMLP) and convolutional neural network. To be specific, this paper first applies these prediction models for stock preselection before portfolio formation. Then, this paper incorporates their predictive results in advancing mean–variance (MV) and omega portfolio optimization models. In order to present the superiority of these models, portfolio models with autoregressive integrated moving average’s return prediction are used as benchmarks. Evaluation is based on historical data of 9 years from 2007 to 2015 of component stocks of China securities 100 index. Experimental results show that MV and omega models with RF return prediction, i.e., RF+MVF and RF+OF, outperform the other models. Further, RF+MVF is superior to RF+OF. Due to the high turnover of these two models, this paper discusses their performance after deducting the transaction fee cased by turnover. Experiments present that RF+MVF still performs the best among MVF models and omega model with SVR prediction (SVR+OF) performs the best among OF models. Moreover, RF+MVF performs better than SVR+OF and high turnover erodes nearly half of their total returns especially for RF+OF and RF+MVF. Therefore, this paper recommends investors to build MVF with RF return prediction for daily trading investment.

Educational timetabling is an ongoing challenging administrative task that is required in most academic institutions. This is mainly due to a large number of constraints and requirements that have to be satisfied. Educational timetabling problems have been classified as NP-hard problems and can be divided into three types: exam timetabling, course timetabling and high school timetabling. The domain of high school timetabling is not well developed when compared to other fields of educational timetabling such as university exam timetabling and course timetabling. As the evolution of the educational systems are continuous, new challenges often arise, requiring new models and solution methodologies. Over the years, a number of methodologies have been developed to address high school timetabling problems. However, there are no comparative studies or rigorous analysis of these methodologies. This survey paper aims to provide a scientific review of high school timetabling. The paper presents a categorisation of the methodologies conducted in recent years based on chronology, category and application (dataset). We first present comparative studies on the success of proposed methodologies. The components and mechanisms of different methodologies are analysed and compared. We also discuss their performance, advantages, disadvantages and potential for improvement. Methodology wise, a shift of popularity from meta-heuristic to mathematical optimisation is observed in recent years. Another observation is that more researchers are opting for XHSTT formatted datasets as a testbed for their algorithms. Finally, we outline the industrial perspective, trends and future direction in high school timetabling optimisation problems.

Conflicting objectives such as minimizing weight and minimizing the maximum nodal displacement, with constraints on normal stresses in the bars, is a common multi-objective structural optimization problem widely found in the literature. This paper proposes multi-objective structural optimization problems with the combination of new conflicting objectives functions and constraints, such as the natural frequencies of vibration and the load factors concerning the global stability of the structure. The solution for these problems may be of great interest in the field of structural engineering, not yet discussed in the literature. The problems analyzed in this paper deal with both discrete and continuous sizing, shape, and layout design variables. The search algorithm adopted here is a modified version of the Differential Evolution called the Third Evolution Step Differential Evolution (GDE3). Several experiments are analyzed with their Pareto-fronts showing the non-dominated solutions. The solutions are defined after obtaining the Pareto curve, which is one of the most important steps and a task that may not be trivial for the Decision Maker. This paper involves a strategy that establishes criteria defining weights (importance) for each objective function and, through these values, enables comparison scenarios. The numerical experiments include plane and spatial benchmark trusses.

Smartphone sensing capabilities have enabled human activity recognition (HAR) solutions to better understand human behavior through computational techniques. However, such solutions have suffered from scalability problems due to the high consumption of computational resources (e.g. memory and processing) and the difficulty of acting in real time due to not observing data evolution over time. These problems occur because the HAR solutions for smartphones have been solved through offline learning with models limited by a data history. The disadvantage of this approach is that human activities constantly change over time and are strongly influenced by the physical environment and user profile. To overcome such problem, this paper proposes a novel low-cost learning algorithm called NOHAR (NOvelty discrete data stream for Human Activity Recognition), focused on continuous flow of data analysis. NOHAR is an online classification algorithm based on symbolic data generated by a discretization process using algorithms as SAX and SFA. The advantages of these algorithms are their abilities to compress and reduce the dimensionality of data. In addition, this paper proposes a new framework called DISTAR (DIscrete STream learning for Activity Recognition) focused on the data streaming analysis. Its goals include standardizing the development of online algorithms for symbolic data. Experimental results using three databases show that NOHAR is on average 33 times faster than the state of the art and can reduce memory consumption by an average of 99.97%.

Differential Evolution Algorithm (DE) is a commonly used stochastic search method for solving real-valued numerical optimization problems. Unfortunately, DE's problem solving success is very sensitive to the internal parameters of the artificial numerical genetic operators (i.e., mutation and crossover operators) used. Although several mutation and crossover methods have been developed for DE, there is not still an analytical method that can be used to select the most efficient mutation and crossover method while solving a problem with DE. Therefore, selection and parameter tuning processes of artificial numerical genetic operators used by DE are based on a trial-and-error process which is time consuming. The development of modern DE versions has been focused on developing fast, structurally simple and efficient genetic operators that are not sensitive to the initial values of their internal parameters. Problem solving successes of the Universal Differential Algorithms (uDE) are not sensitive to the structure and internal parameters of the related artificial numerical genetic operators used, unlike DE. In this paper a new uDE, Bezier Search Differential Evolution Algorithm, BeSD, has been proposed. BeSD’s mutation and crossover operators are structurally simple, fast, unique and produce highly efficient trial patterns. BeSD utilizes a partially elitist unique mutation operator and a unique crossover operator. In this paper, the experiments were performed by using the 30 benchmark problems of CEC2014 with Dim=30, and one 3D viewshed problem as a real world application. The problem solving success of BeSD was statistically compared with five top-methods of CEC2014, i.e., CRMLSP, MVO, WA, SHADE and LSHADE by using Wilcoxon Signed Rank test. Statistical results exposed that BeSD’s problem solving success is better than those of the comparison methods in general.

Feature selection is crucial to develop a brain computer interface (BCI) system which has high classification accuracy and less computational complexity in especially a large feature space. Feature selection (FS) problem has been solved by many various methods. Among these methods, especially evolutionary computation (EC) techniques have gained a lot of attention in recent years. However, there are very few studies in the literature that consider FS problem as a multi-objective problem to find the optimal trade-off between classification accuracy and the number of selected features. Therefore, in this paper, a non-dominated sorting multi-objective symbiotic organism search (NSMOSOS) algorithm is proposed to generate the optimal feature subset in BCI. The efficiency and robustness of the proposed algorithm as a feature selection method is investigated in two datasets based on motor imagery. The highest classification accuracies of NSMOSOS for dataset 1 and dataset 2 are obtained 97.86% with 11 features and 96.57% with average 19 features, respectively. The obtained results demonstrate that the proposed method achieves satisfying results with regard to both the classification accuracy improvement and feature reduction rates for both datasets. The superiority of the proposed method is verified compared with the existing methods for both datasets. Besides, three different versions of symbiotic search organism (SOS) algorithm are improved, and pros and cons of these algorithms are evaluated compared with each other. In conclusion, the paper indicates that the proposed NSMOSOS algorithm is an efficient and practicable technique for FS problem and could be helpful in developing the BCI applications.

Finding feasible solutions to real-world problems is a crucial task. Metaheuristic algorithms are widely used in many fields due to the variety of solutions they can produce. The grey wolf optimizer (GWO) is a relatively novel population-based metaheuristic algorithm that has been shown to have good optimization performance. However, due to the insufficient diversity of wolves in some cases, this approach can lead to locally optimal situations. Therefore, this paper proposes a grey wolf optimization method based on a beetle antenna strategy (BGWO) that gives the leader wolf a sense of hearing to improve the global search ability and reduce unnecessary searches. In addition, to balance exploration and exploitation, a nonlinear dynamic control parameter update strategy based on the cosine function is proposed. To evaluate the performance of the proposed BGWO, this paper uses 23 standard benchmark functions to test the method in different dimensions. Moreover, four well-known engineering problems are used to evaluate the ability of the proposed algorithm to obtain real-world problem solutions. The experimental results show that BGWO has superior performance and is competitive with many state-of-the-art algorithms in terms of solution accuracy, convergence rate, and stability.

Proteomics is nowadays one of the most important and relevant fields from computational biology, raising a lot of challenging and provocative questions. Gaining an understanding of protein dynamic and function as well as obtaining additional insights into the protein folding process is still of great interest in bioinformatics and medicine. This paper introduces a new approach A n o m a l P for detecting anomalous protein conformational transitions using deep autoencoders for encoding information about the structural similarity between proteins belonging to the same superfamily. Experiments are conducted on real protein data and the obtained results emphasize the potential of autoencoders to learn biological relevant patterns, such as proteins’ structural characteristics and that they are useful for detecting conformations or proteins which are likely to be anomalous with respect to a superfamily. The study performed in this paper is aimed to provide better insights of proteins structural similarity, with the broader goal of learning to predict proteins conformational transitions.

This paper analyses different multi-objective evolutionary algorithms to deal with the Environmental/Economic Load Dispatch (EELD). EELD is formulated as a multi-objective optimization problem in which two competing objectives (fuel cost and pollutants emission) should be optimized simultaneously while fulfilling constraints. Due to the typical process of an evolutionary algorithm (EA), the use of operators applied to individuals of the population might violate constraint rules of the problem. The way in which EAs deal with such constraint rules is an important point and it is directly related to the quality of the generated solutions. One of the contributions of this paper is the analysis of the impact of a repair procedure in four multi-objective EAs. The analyzed approaches are evaluated in eight known instances (with 3, 6, 10, 20 and 40 generators) of the multi-objective EELD. Furthermore, two new instances (with 80 and 120 generators) are proposed and evaluated in this work. Experiments were applied using Dominance Ranking, hypervolume and unary-epsilon indicators, empirical attainment functions and statistical tests, in order to evaluate the algorithms performances. The results point to the consistency of the NSGA-II with repair procedure compared to the literature algorithms, and it outperforms other approaches in most of the considered instances.

The excellent generalization performance of time series ensemble forecasting depends on the accuracy and diversity of the individual models. In this paper, a heterogeneous ensemble forecasting model with multi-objective programming for nonlinear time series is proposed. Accordingly, an improved multi-objective particle swarm optimization (MOPSO) algorithm integrated with a dynamic heterogeneous mutation operator is designed. The nonlinear time series of the Baltic Dry Index (BDI) is selected as the forecasting object to train, validate and test the ensemble forecasting model established in this paper. To verify the superior forecasting performance of the proposed model, 20 forecasting models including statistical models, machine learning models, and optimization algorithm–based ensemble models are utilized and compared. The experimental results under different lead times revealed that: 1) the forecasting approach with multi-objective programming has excellent robustness and can effectively exert out-of-sample prediction under different lead times for nonlinear time series; 2) with the increase of lead time, the out-of-sample forecasting performance would gradually decrease for all models, and the precision of the ensemble forecasting model is better than that of the individual forecasting model; 3) the forecasting performance of the MOPSO with crowding distance (MOPSOCD)-based ensemble forecasting model is better than that of benchmark machine learning models and other optimal ensemble forecasting models in terms of the prediction accuracy and statistical test results.

Many intelligent systems in business, government or academy process natural language as an input during inference or they might even communicate with users in natural language. The natural language processing is currently often done with machine learning models. However, machine learning needs training data and such data are often scarce for low-resource languages. The lack of data and resulting poor performance of natural language processing can be solved with cross-lingual learning. Cross-lingual learning is a paradigm for transferring knowledge from one natural language to another. The transfer of knowledge can help us overcome the lack of data in the target languages and create intelligent systems and machine learning models for languages, where it was not possible previously. Despite its increasing popularity and potential, no comprehensive survey on cross-lingual learning was conducted so far. We survey 173 text processing cross-lingual learning papers and examine tasks, datasets and languages that were used. The most important contribution of our work is that we identify and analyze four types of cross-lingual transfer based on “what” is being transferred. Such insight might help other NLP researchers and practitioners to understand how to use cross-lingual learning for wide range of problems. In addition, we identify what we consider to be the most important research directions that might help the community to focus their future work in cross-lingual learning. We present a comprehensive table of all the surveyed papers with various data related to the cross-lingual learning techniques they use. The table can be used to find relevant papers and compare the approaches to cross-lingual learning. To the best of our knowledge, no survey of cross-lingual text processing techniques was done in this scope before.

Social media has recently become popular due to its vast applications. The common people all over the world uses its diverse channels to express personal views, experiences and opinions regarding diverse topics. Social media has revolutionized the way people interact and communicate with each other and overall, it has changed the methods and approaches in about all the aspects of life such as social issue, business, education, health, etc. Thus, sales and marketing departments of multinational industries are focusing on social media trends to analyze current trends and predict future trends by analyzing user generated content on Facebook, Flickr, Twitter, etc. However, the prediction process becomes challenging as the multiplicity of factors affect the popular elements in the social media content. This research paper aims to work on Twitter trend analysis and proposes a trend detection process over streams of tweets. The proposed approach detects the trending topics of the real-time Twitter trends along with ranking the top terms and hashtags. The paper further discusses the motivation for trend prediction over the social media; In addition to exploratory data analysis, the research paper explores the Term Frequency-Inverse Document Frequency (Tf-IDF), Combined Component Approach (CCA) and Biterm Topic Model (BTM) approaches for finding the topics and terms within given topics. In modern competitive world, this research provides investors, advertisers, industries and all the stakeholders. A detailed and comprehensive data analysis which may help them to focus their investment, area of work, marketing, and product.

Pythagorean fuzzy (PF) sets involving Pythagorean membership grades can befittingly manipulate inexact and equivocal information in real-life problems involving multiple criteria decision analysis (MCDA). The linear programming technique for multidimensional analysis of preference (LINMAP) is a prototypical compromising model, and it is widely used to carry on decision-making problems in many down-to-earth applications. In LINMAP, the employment of squares of Euclidean distances is a significant technique that is an effective approach to fit measurements. Taking the advantages of a newly developed Euclidean distance model on the grounds of PF sets, this paper initiates a beneficial concept of squared PF Euclidean distances and studies its valuable and desirable properties. This paper aims to establish a squared Euclidean distance (SED)-based outranking approach and develop a novel PF LINMAP methodology for handling an MCDA problem under PF uncertainty. In the SED-based outranking approach, a novel SED-based dominance index is proposed to reflect an overall balance of a PF evaluative rating between the connection and remotest connection with positive- and negative-ideal ratings, respectively. The properties of the proposed index are also analyzed to exhibit its efficaciousness in determining the dominance relations for intracriterion comparisons. Moreover, this paper derives the comprehensive dominance index to determine the overall dominance relation and defines measurements of rank consistency for goodness of fit and rank inconsistency for poorness of fit. The PF LINMAP model is formulated to seek to ascertain the optimal weight vector that maximizes the total comprehensive dominance index and minimizes the poorness of fit under consideration of the lowest acceptable level and specialized degenerate weighting issues. The practical application concerning bridge-superstructure construction methods is conducted to test the feasibility and practicability of the PF LINMAP model. Over and above that, a generalization of the proposed methodology, along with applications to green supplier selection and railway project investment, is investigated to deal with group decision-making issues. Several comparative studies are implemented to further validate its usefulness and advantages. The application and comparison results display the effectuality and flexibility of the developed PF LINMAP methodology. In the end, the directions for future research of this work are represented in the conclusion.

With the rising share of electric vehicles used in the service industry, the optimization of their specific constraints is gaining importance. Lowering energy consumption, time of charging and the strain on the electric grid are just some of the issues that must be tackled, to ensure a cleaner and more efficient industry. This paper presents a Two-Layer Genetic Algorithm (TLGA) for solving the capacitated Multi-Depot Vehicle Routing Problem with Time Windows (MDVRPTW) and Electric Vehicles (EV) with partial nonlinear recharging times (NL) – E-MDVRPTW-NL. Here, the optimization goal is to minimize driving times, number of stops at electric charging stations and time of recharging while taking the nonlinear recharging times into account. This routing problem closes the gap between electric vehicle routing problem research on the one hand and its applications to several problems with numerous real-world constraints of electric vehicles on the other. Next to the definition and the formulation of the E-MDVRPTW-NL, this paper presents the evolutionary method for solving this problem using the Genetic Algorithm (GA), where a novel two-layer genotype with multiple crossover operators is considered. This allows the GA to not only solve the order of the routes but also the visits to electric charging stations and the electric battery recharging times. Various settings of the proposed method are presented, tested and compared to competing meta-heuristics using well-known benchmarks with the addition of charging stations.

Several algorithms have been proposed towards discovering the graphical structure of Bayesian networks. Most of these algorithms are restricted to observational data and some enable us to incorporate knowledge as constraints in terms of what can and cannot be discovered by an algorithm. A common type of such knowledge involves the temporal order of the variables in the data. For example, knowledge that event B occurs after observing A and hence, the constraint that B cannot cause A . This paper investigates real-world case studies that incorporate interesting properties of objective temporal variable order, and the impact these temporal constraints have on the learnt graph. The results show that most of the learnt graphs are subject to major modifications after incorporating incomplete temporal objective information. Because temporal information is widely viewed as a form of knowledge that is subjective, rather than as a form of data that tends to be objective, it is generally disregarded and reduced to an optional piece of information that only few of the structure learning algorithms may consider. The paper argues that objective temporal information should form part of observational data, to reduce the risk of disregarding such information when available and to encourage its reusability across related studies.

Indoor localization systems have the capability to change the way of providing location-based services in a closed environment. Though there is no agreed-upon technology that works best in indoor, WiFi signal is an important alternative as most of such places are covered by WiFi Access Points (APs). In this paper, the problem of indoor localization is investigated from the perspective of expert systems through applying machine learning techniques. The significant variation of WiFi signal strength with ambient conditions as well as device configuration badly affects the localization accuracy. Thus, the fingerprinting effort required to train a localization system subject to context heterogeneity is huge. The uncertainty in localization performance due to varying contexts is hardly investigated in the literature. Consequently, the main contribution of this paper is to propose a weighted ensemble classifier based on Dempster–Shafer belief theory to efficiently handle context heterogeneity. Here, the context is defined in terms of different smartphone configurations used for training and testing the system as well as temporal variation of signals. The method presented here utilizes the Dempster–Shafer theory of belief functions to calculate the weights of the base learners in the decision of the ensemble. Belief theory is applied here to handle the inherent uncertainty in WiFi signal variations due to heterogeneous context. Real life experiments are conducted for two datasets, JUIndoorLoc and UJIIndoorLoc at different granularity levels. For JUIndoorLoc, with state-of-the-art classifiers, 86–97% accuracy can be achieved for 10-fold cross-validation. However, when the training context differs from the test conditions, accuracy drops to 62–87%. In such a scenario, the proposed weighted ensemble technique is found to achieve almost 98% localization accuracy when RSSIs, mean and variance of RSSIs are considered as features. The technique can lead to an effective expert system for indoor localization at varying granularity levels. Such systems would be beneficial for pervasive indoor positioning applications as no dedicated infrastructure is needed for positioning.

This paper examines the option pricing performance of the most popular Machine Learning algorithms. The classic parametrical models suffer from several limitations in term of computational power required for parametric calibration and unrealistic economical and statistical assumptions. Therefore, a data driven approach based on non-parametric models is are well justified. Most of the previous researchers focus especially on the neural networks method (NN), the other algorithms being unexplored. Beside NN, this paper also analyses the performance of the Support Vector Regressions and Genetic Algorithms and propose three other Decision Tree methods, respectively Random Forest, XGBoost and LightGMB. In order to emphasize the power of this algorithms, a comparison with classical methods like Black-Scholes and Corrado-Su with both historical and implied parameters have been conducted. The analyzes were performed on European call options who have as underlying asset the WTI crude oil future contracts. Machine Learning algorithms outperform by a great margin the classical approaches regardless of the moneyness and the maturity of the contracts.

Decentralized nature of cryptocurrencies, irrational cryptocurrency valuations and severe price volatility in cryptocurrency market makes it a formidable task for investors to pick individual coins, and rather investors would prefer to invest money on the entire cryptocurrency market accurately represented by a cryptocurrency index. This paper proposes the design of a Principal Component Analysis based methodology to construct a dynamic cryptocurrency index that accurately tracks the movement of the entire cryptocurrency market. Our analysis on real market data shows why first component derived from PCA is sufficient to construct the cryptocurrency index and how to determine the number of constituents while building the index. Proposed PCA based tool, tested on actual historical cryptocurrency data, takes into account the changing dynamics of the cryptocurrency market by regularly shifting the number of constituents as well as the weights and is able to capture the evolving pattern in the cryptocurrency market. The proposed index has been validated by three-factor pricing model, consisting of market, size and momentum factors. Subsequently, the proposed index has been compared with other existing indexes based on the results of three-factor model. In summary, the paper presents a robust mathematical model for construction of a dynamic cryptocurrency index that can be used as a tool to analyze the return on investments as well as to study the fluctuations present in the cryptocurrency market.