Search

In this paper we present a new procedure for patent selection and ranking that uses a new dynamic consensus model and a new variant of the Matching Method real options framework. The procedure utilizes opinions from multiple experts in the creation of a stream of consensual yearly cash-flows for each patent. The cash-flow stream is used as an input in continuous-time real option valuation of each patent. Continuous time analysis of the patents allows for optimizing the timing of dropping patents from the portfolio. The procedure is relatively easy to apply in a practical setting, and is consistent with financial theory. Using the procedure is illustrated with a numerical example. Ranking result sensitivity to parameter value selection is studied separately.

In this paper, we propose a novel hybrid fuzzy–metaheuristic approach with the aim of overcoming premature convergence when solving multimodal single and multi-objective optimization problems. The metaheuristic algorithm used in our proposed approach is based on the imperialist competitive algorithm (ICA), a population-based method for optimization. The ICA divides its population into sub-populations, known as empires. Each empire is composed of a high fitness solution—the imperialist—and some lower fitness solutions—the colonies. Colonies move towards their associated imperialist in each empire to achieve better status (higher fitness). The most powerful empire tends to attract weaker colonies. These competitions and movements can be enhanced for better algorithm performance. In our hybrid approach, a global learning strategy is devised for each colony to learn from its best-known position, its associated imperialist and the global best imperialist. A fast-evolutionary elitism local search is used to enhance the collaborative search mechanism (competition) in each empire, and thus the overall optimization performance may be improved. Other main evolutionary operators include velocity adaptation and velocity divergence. To address parameterization and computational cost evaluation issues, two fuzzy inferencing mechanisms are designed and used in parallel: one is a learning strategy adaptor in each run, and the other is a smart evolution selector in each running window. For Pareto front approximation, fast-elitism non-dominated sorting is applied to the solutions, and a novel penalized sigma diversity index is designed to estimate the diversity (power) of solutions in the same rank. Comprehensive experimental results based on 22 well-known single-objective and 25 multi-objective benchmark instances clearly show that our proposed approach has better solution quality compared with other popular metaheuristics and state-of-the-art ICA variants. The proposed approach can be used as an optimization module in any intelligent decision-making systems to enhance efficiency and accuracy. The designed fuzzy inferencing mechanisms can also be incorporated into any single- or multi-objective optimizers for parameter tuning purposes, to make the optimizers more adaptive to new problems or environments.

In this paper, we propose a novel density-based radar scanning clustering algorithm. Its main objective is to quickly discover and accurately extract individual clusters by employing the radar scanning strategy. By using this algorithm, the number of clusters does not need to be specified beforehand. Two techniques are utilized in our proposed method. First, we use a fast mean-shift algorithm with adaptive radius and active subsets to effectively locate the centers, reducing the computational time significantly. Second, we employ the shape of the probability density function of the distribution of distances between a selected point and the other points in the data set. This is performed to determine the critical parameters of the radiuses of the fast mean-shift algorithm and radiuses of clusters. The new algorithm has four merits. It reduces the computational complexity, overcomes problems caused by high dimensionality, is capable of dealing with heterogeneous spherical data sets, and lastly, is robust to noise and outliers. After applying our proposed method to several kinds of synthetic and real-world data sets, the results indicate that the density-based radar scanning algorithm is efficient and accurate.

In this paper, we present a fraud detection framework for publicly traded firms using an optimization approach integrated with a meta-heuristic algorithm known as Beetle Antennae Search (BAS). Existing techniques include human resources, like financial experts and audit teams, to determine the ambiguities or financial frauds in the companies based on financial and non-financial ratios. It is a laborious task, time-consuming, and prone to errors. We designed an optimization problem to minimize the loss function based on a non-linear decision function combined with the maximization of recall (Sensitivity and Specificity). We solved the optimization problem iteratively using the BAS. It is a nature-inspired algorithm and mimics the beetle’s food-searching nature. It includes a single searching particle to find an optimal solution to the optimization problem in n -dimensional space. We used a benchmark dataset collected from SEC’s Accounting and Auditing Enforcement Releases (AAERs) for the simulation. It includes 28 raw financial variables and the data collected between 1991- 2008 . For the comparison, we evaluated the performance of BAS with the recently proposed approach using RUSBoost. We also compared it with some additional algorithms, i.e., Logit and SVM-FK. The results showed that BAS is comparable with these algorithms and outperformed them in time consumption.

Nowadays, in scientific and computational programs, increasing the execution speed of programs is very important. In implementing scientific programs, loops are dedicating a large part of time to themselves. Therefore, the parallel execution of the iteration loop reduces the runtime of all programs. This paper has been working on the parallelism of two-level perfectly nested loops with non-uniform dependencies and rectangular iteration space, and then their scheduling is based on the scheduling of task graphs. For this purpose, the iteration space of non-uniform loop is partitioned using a tiling transformation, and then the tiled space is modelled as a task, graph for scheduling. In the task graph, vertices expressive tiles and edges expressive data dependencies inter-tiles. It should be noted that the computational load of vertices is not the same, and the vertices corresponding imperfect tiles have a less computational load. The scheduling is done on both types of homogeneous and heterogeneous multiprocessor systems. The tiling of iteration space loop and the scheduling of task graph is arranged in the category of NP-Hard problems that using evolutionary algorithms can be efficient to discover a suitable solution for them. The results of the proposed approach show the performance of this method better than previous methods in the tiling and scheduling of non-uniform loops. According to the examinations performed on uniform loops, this algorithm also performs better than previous algorithms on these types of loops.

Real-world data are often corrupted by noise and outliers, which are originated from different procedures such as data collection, storage, and processing. Noise and outliers decrease the quality of clustering and lead to the inaccurate and misplaced cluster centers. In this paper, we propose a new algorithm called Improved Possibilistic Fuzzy C-Means (IPFCM) to cluster noisy data. First, initial cluster centers are calculated by Possibilistic Fuzzy C-Means (PFCM) which do not match dense regions of the data. Then, the domain is divided to some subdomains and each data point is assigned to a sub-domain. The cluster centers are iteratively moved towards high-density regions by maximizing a novel cluster validity index. In the proposed method, a Gaussian membership function is defined on each cluster to weight the data. Then, the sum of weights in each cluster is calculated. The product of these values is considered as the validity index. Since division of the domain is changed with moving the cluster centers, this procedure is repeated until the convergent criterion is satisfied. Cluster analysis performed on six synthetics, nine real benchmarks datasets shows the superiority of IPFCM over some previous clustering algorithms such as Fuzzy C-Means (FCM), PFCM, Kernel Fuzzy C-Means (KFCM), Noise Clustering (NC), and Generalized Entropy based Possibilistic Fuzzy C-Means (GEPFCM). The clustering results of near-fault ground motion data indicate that the cluster centers identified by IPFCM are well separated from each other, while those for PFCM are close to each other in some datasets. Moreover, the results show that the impact of noisy data on the proposed index and consequently cluster analysis decreases as the noisy data get away from the cluster centers which is one of the advantages of using IPFCM algorithm.

This paper proposes a novel nature-inspired meta-heuristic optimizer, called Reptile Search Algorithm (RSA), motivated by the hunting behaviour of Crocodiles. Two main steps of Crocodile behaviour are implemented, such as encircling, which is performed by high walking or belly walking, and hunting, which is performed by hunting coordination or hunting cooperation. The mentioned search methods of the proposed RSA are unique compared to other existing algorithms. The performance of the proposed RSA is evaluated using twenty-three classical test functions, thirty CEC2017 test functions, ten CEC2019 test functions, and seven real-world engineering problems. The obtained results of the proposed RSA are compared to various existing optimization algorithms in the literature. The results of the tested three benchmark functions revealed that the proposed RSA achieved better results than the other competitive optimization algorithms. The results of the Friedman ranking test proved that the RSA is a significantly superior method than other comparative methods. Finally, the results of the examined engineering problems showed that the RSA obtained better results compared to other various methods.

The elicitation of expert knowledge in a Bayesian model is a big challenge. Elicitation requirements increase exponentially with the size of the model, making the expert’s task more difficult and generating problems of quality or coherence of the elicited result. In this paper, we develop an extended Cluster-WSA method using several clusters as an elicitation method. A general formula to calculate the number of information to elicit and a generic algorithm for this method are developed. A selection procedure of some elicitation methods is established and the usefulness of the Cluster-WSA method is verified. The case study of parcel delivery with a drone is used to apply our method, to build a decision support system for diagnosis and prediction, and to make a performance indicator system.

Nowadays, cloud customers use cloud services increasingly to satisfy their demands. Usually, a significant number of customers are immature and inexpert and cannot express their needs accurately and numerically. They usually express their needs verbally and in the form of linguistic terms. On the other hand, the experienced customers express their needs numerically and accurately. In this situation, a recommendation system can be considered as one of the most useful ideas to support all type of customers. However, current recommendation systems (e.g., collaborative filtering based recommendations) meet customer requests that are accurately and numerically expressed. To support all types of customers, the construction of a strong recommendation system to analysis the demands expressed by customers (experienced and inexperienced) and to recommend suitable services is vital. As another important matter, cloud customers and services have been geographically distributed. Identifying the location of customers and services has a significant effect on the quality of services offered to customers. Therefore, the recommendation system should consider the location of customers and services in order to provide better services. In this paper, we introduce an efficient method to construct a powerful recommendation system which can provide suitable services considering the preferences of the customer and their location. The proposed recommendation system comprises two algorithms. The first algorithm is a fuzzy clustering algorithm, named FCA, that can well classify the location of customers and services. The second algorithm is an iterative adaptive neural-fuzzy algorithm, named IANFRA, which receives the preferences of the customer along with their location and identifies suitable services based on the locations clustered by FCA and the demands of customers (experienced and inexperienced). Finally, the feasibility of the proposed method has validated in terms of accuracy and scalability through conducting extensive experiments on a real distributed service quality dataset WS-DREAM. The evaluation results illustrate that both the service recommendation accuracy in the prediction of quality of services and the scalability, when the volume of the dataset is huge, have been improved.

Every music composition has a composer at the core of its building block, molding it into a style of their own. The creative compositional style of a composer varies dynamically with every composer which is perishable and inimitable but cannot be preserved. This paper proposes a hybrid style transfer model which is an end-to-end approach to transfer style, compose as well as evaluate the style transfer accuracy in the domain of the target music composer. The paper focuses on 3 composer maestros Liszt, Chopin and Schubert taken from the Maestro dataset. The main step of music composer style transfer is accomplished in 3 systematic steps, training composer classifiers according to feature sets, generating a model of a composition in a particular composer’s style from noise, and finally style transfer and its evaluation. The 3 steps are interconnected in a way that each step is the building block for the next step. The GAN architecture of the style transfer step is built out of the GAN architecture of the second step. The compositions generated from each architecture is finally evaluated by the common pre-trained classifier of the first step. The highest accuracy obtained is 80% for composer classification using the maestro dataset, 77.27% for the classification of the generated style transferred version of a composition into the target composer class using the pre-trained classifiers.

In this paper, we propose a novel framework for transferred deep learning-based anomaly detection in hyperspectral images. The proposed framework includes four main steps. Firstly, the image2019s spectral dimension is reduced by applying the principal component analysis (PCA) to decrease computational time. Secondly, a deep convolutional neural network (CNN) is trained using only one image to learn the pixels2019 similarities in a picture. Consequently, a novel and well-designed algorithm entitled object area filtering (OAF) is employed to benefit from this learned similarity for extracting objects in the image. The OAF removes irrelevant objects by comparing their area to an acceptable anomaly area range. Lastly, the final result is obtained by multiplying the network output and binary map of anomalies. The receiver operating characteristic (ROC) is employed to evaluate the proposed framework. Extensive experimental evaluations demonstrate that the proposed framework substantially outperforms a significant number of comparable state-of-the-art methods. Finally, we empirically verify that the deep network exhibits excellent domain adaptability.

Satellite images have become more easily accessible due to government programs and the availability of commercial earth observation satellites. The increased volume of data and the recent advancement in deep learning-based techniques have made possible the development of fast and accurate methods for satellite image segmentation, which aims at automatically giving each pixel in the satellite image a label of the underlying object/ground type. However, despite the abundant supply of raw satellite images, the labelling of these images is a manual task and is therefore time and labour intensive. When a large labelled dataset is absent, resorting to a smaller data source significantly hampers the performance of the machine learning model. Therefore, an approach which mitigates the effects of using small labelled datasets should be investigated. This paper provides a solution to the problem of insufficient data samples for satellite image segmentation. To this end, two methods are developed and compared with the benchmark U-Net model: the use of transfer learning on an Xception architecture, and the use of steerable filters. The first model transfers knowledge from a pre-trained Xception model and predicts with a multi-resolution feature fusion module, specifically designed to recover fine details. The latter further improves the capability of our neural network model in handling data with high rotational variation. Both models are experimentally shown to significantly enhance the segmentation performance despite a small-sized training dataset.

We introduce a multidimensional multiblock clustering (MDMBC) algorithm in this paper. MDMBC can generate overlapping clusters with similar values along clusters of dimensions. The parsimonious binary vector representation of multidimensional clusters lends itself to the application of efficient meta-heuristic optimization algorithms. In this paper, a hill-climbing (HC) greedy search algorithm has been presented that can be extended by several stochastic and population-based meta-heuristic frameworks. The benefits of the algorithm are demonstrated in a bi-clustering benchmark problem and in the analysis of the Leiden higher education ranking system, which measures the scientific performance of 903 institutions along four dimensions of 20 indicators representing publication output and collaboration in different scientific fields and time periods.

Mining frequent closed itemsets over data streams is an important data mining problem. Mining data streams is more challenging than mining static data because of the nature of data streams, including high arrival rate, massive volume of incoming data, and concept drift. The existing algorithms for mining frequent closed itemsets over data streams suffer from scalability and efficiency bottlenecks. This paper proposes a novel algorithm for mining frequent closed itemsets over data streams both for the sliding window model and for the landmark model. An indexed prefix closed itemset tree is proposed for compressing all closed itemsets and for quick searching of closed itemsets, and novel search strategies are proposed to prune the search space in updating the set of closed itemsets. The proposed algorithm outperforms the state-of-the-art intersection-based algorithms, CICLAD, ConPatSet, and CloStream, by several times to 2 orders of magnitude in efficiency, and also outperforms the state-of-the-art pattern enumeration algorithm, Moment, by up to 2 orders of magnitude over data streams with large windows and sparse data streams. The proposed algorithm is also superior in scalability.

The paper aims at introducing novel distance measures for the intuitionistic fuzzy set (IFS) to discuss the decision-making problems. The current work exploits four different notions of centers, namely centroid, orthocenter, circumcenter and incenter of the triangle. First, we mold knowledge embedded in IFSs into isosceles TFN (triangular fuzzy number). Hence, based on these TFNs, we design four-novel distance/similarity measures for IFSs using the structures of the four aforementioned centers and inspect their properties. To avoid the loss of information during the conversion of IFSs into isosceles TFNs, we included the degree of hesitation (t) between the pairs of the membership function in the process. The compensations and authentication of the proposed measures are established with diverse counter-intuitive patterns and decision-making obstacles. Further, a clustering algorithm is also given to match the objects based on confidence levels. The performed analysis shows that the proposed measures give distinguishable and compatible results as contrasted to existing ones.

There have been many studies on machine learning and data mining algorithms to improve the effect of credit risk assessment. However, there are few methods that can meet its universal and efficient characteristics. This paper proposes a new multi-classification assessment model of personal credit risk based on the theory of information fusion (MIFCA) by using six machine learning algorithms. The MIFCA model can simultaneously integrate the advantages of multiple classifiers and reduce the interference of uncertain information. In order to verify the MIFCA model, dataset collected from a real data set of commercial bank in China. Experimental results show that MIFCA model has two outstanding points in various assessment criteria. One is that it has higher accuracy for multi-classification assessment, and the other is that it is suitable for various risk assessments and has universal applicability. In addition, the results of this research can also provide references for banks and other financial institutions to strengthen their risk prevention and control capabilities, improve their credit risk identification capabilities, and avoid financial losses.

In this paper, the bi-objective quadratic multiple knapsack problem is tackled with a hybrid population-based method. The proposed method starts by computing two reference solutions, where a specialized powerful mono-objective algorithm is used. From both reference solutions, a starting population is built by using a series of perturbations around the solutions. Next, the so-called non-sorting genetic process is combined with a new drop/rebuild operator for generating a series of populations till converging toward an approximate Pareto front with high density. The performance of the hybrid population based algorithm (namely HBPA) is evaluated on a set of benchmark instances of the literature containing both medium and large-scale instances. Its provided results are compared to those achieved by the best methods available in the literature. Encouraging results have been obtained.

Forecasting the box offices of movies is indispensable, whether it is for the stock investment decisions of the movie company, the movie sales strategy design of the advertising company, or the movie screening schedules of cinemas. However, this task is very challenging because it is affected by many complicated factors. In view of this complexity, we propose a method for forecasting the box offices of movies coming soon based on social media analysis. First, movie-related online reviews and their numbers are collected from microblogs. Second, several important features, i.e., the sentiment value of microblog texts, the consumers’ RFM value and the degree of moviegoers’ attention, are extracted or calculated. Subsequently, we use the extracted features to expand the parameters of the Bass model, and build two improved Bass models for forecasting the box offices of movies in the initial week and during the diffusion cycle, respectively. Finally, through fitting data progressively, we realize the forecasting task. The experimental results show that compared with the traditional forecast models and existing studies, the improved Bass method proposed in this paper has better performance, overall.

Intuitionistic fuzzy set (IFS), a classical extension of fuzzy set, is a powerful tool to describe fuzzy information in multi-criteria decision-making (MCDM) because of its membership and non-membership functions. Intuitionistic fuzzy aggregation operators are the standard mathematical tools for the combination of several inputs with respect to criteria into one unique output. The traditional intuitionistic fuzzy aggregation operator-based methods have been criticized because of reasons that include disregard for comprehensive correlative relationships of criteria and ignorance for combination of fuzzy measures of criteria and weights of positions. In this paper, the effectiveness of traditional aggregation operator-based MCDM techniques is improved by a novel method to select the best alternative(s) under the intuitionistic fuzzy environment. We begin by defining the Choquet integral-based intuitionistic fuzzy arithmetic aggregation (CIIFAA) operator, the remarkable properties of which are proved in details. Further, the Choquet integral based intuitionistic fuzzy hybrid arithmetic aggregation (CIIFHAA) operator is defined as well as the proofs of its properties. The CIIFHAA operator can not only capture the comprehensive correlative relationships of criteria in a simpler manner, but also combine the weights of positions. Then, we put forward a MCDM method based on CIIFHAA operator with intuitionistic fuzzy evaluations. An illustrative example is solved, the best alternative and the ranking of alternatives are obtained, which can verify the feasibility of the proposed method. Finally, two comparisons are conducted to illustrate the stability and advantages of the proposed method.

The prediction of stock price has always been a main challenge. The time series of stock price tends to exhibit very strong nonlinear characteristics. In recent years, with the rapid development of deep learning, the ability to automatically extract nonlinear features has significantly attracted scholars’ attention. However, the majority of the relevant studies have concentrated on prediction of the changes of stock market based on the data of the specific stock (e.g., transaction data, financial data, etc.), while those studies ignored the interaction between stocks of different industries, especially the interaction between the stocks of upstream enterprises and downstream enterprises in the industrial chain. This paper aims to propose a combination of transfer learning of industrial chain information and deep learning models, including multilayer perceptron (MLP), recurrent neural network (RNN), long short-term memory (LSTM), and gated recurrent unit (GRU), for stock market prediction. These models are used to predict the trend of the 379 stock market indices by industry in China, and the DM test was employed for validation of the prediction results. It can be concluded that RNNs are not necessarily such an optimal choice for the prediction when dealing with specific time series data, and it could be justified by using the local interpretable model-agnostic explanations (LIME) algorithm. Hence, the MLP was selected to effectively improve the accuracy of the prediction of the stock market indices based on the transfer learning of industrial chain information. The investment strategy is constructed according to the prediction results, and the yield of maturity exceeds that of the buy-and-hold strategy.

Supplier evaluation is a complex multi-criteria decision-making (MCDM) problem that deals with assessment of suppliers as the potential alternatives against various types of criteria. We consider the context where decision makers (DMs) have complete information about the suppliers and criteria. To address the needs of decision makers, a multi-criteria evaluation method named Ranking based on optimal points (RBOP) is developed in this paper. By imitating and simulating human decision-making behavioural patterns, the developed MCDM method selects the best alternative that is closer to what the DM desires. Furthermore, a novel subjective MCDM weighting method s called win-loss-draw (WLD) method is also developed, which is also based on human behavioural pattern. A real case study of domestic cheese brands is considered to apply the developed methods to select the best cheese supplier for an Iranian hypermarket. Compared to other MCDM methods, outputs of the RBOP method show some differences due to the impact of WLD method, which intensified divergence and optimal points during the decision-making process.

Recommendation systems rely on the historic data of users' purchases and their feedbacks to profile their preferences and make future recommendations. Most of these systems usually employ Collaborative Filtering (CF) models to analyze users’ ratings and infer the latent factors which represent the user and item features in k-dimensional latent space. However, the historical rating data used for recommendations are usually sparsed and unbalanced. Various approaches have been used to resolve these issues by combining the user’s ratings and reviews to better capture the user’s sentiments and make accurate recommendations. Other challenges comprise changes in users’ preferences and items’ perceptions over time. Therefore, this paper presents a new Sentiment Scoring Model (SSM) based on Long-/Short-Term Memory and a combination function that catches the sentiment bias between user rating and review to relieve the sparsity and unbalanced dataset. Next, we proposed an Adaptive LSTM (ALSTM) method that can model the drifting of user and item features to improve the recommendation accuracy. We show the performance of our model on the three real-world rating datasets from Amazon reviews, which comprises Fine Food, Baby, and Cell-phone & Accessories categories. The result shows the superiority of our proposed model over the existing static and dynamic models. The statistical test shows that all the performance gains are significant at p

Deep learning tasks rely heavily on a large amount of training data, but collecting and annotating data daily is not practical. Therefore, Zero-shot learning (ZSL) has become important for the applications, where no labeled data is available during training. ZSL aims at recognizing unseen classes by semantic transfer of information from seen to unseen classes. In this paper, we have proposed an identifiable VAE (iVAE) based generative model to address conventional and generalized ZSL. The key to our approach is learning disentangled representations, where each dimension is statistically independent and responsible for generating data. Thus, VAE is a commonly used model for learning disentangled independent factors of variation from the data. Our goal is to learn a latent space representing significant information, that approximates the actual data distribution. Extensive experiments on five benchmark datasets, i.e. CUB, AWA1, AWA2, SUN and aPY, are performed for further evaluation in both settings.

Thermal cameras work well in harsh environments, but the quality of infrared images is not as high as visible light. Thermal to visible image translation can get rid of the image modal differences caused by various spectral characteristics. Nowadays, Generative Adversarial Network (GAN) can transform the images from one domain to another domain, but the generated images are still in a single channel in case of facial thermal to visible. In this paper, we propose a claw connection-based generative adversarial networks framework named ClawGAN for the facial thermal images to RGB visible images translation. We proposed the mismatch metric ( MM ) to measure the mapping relationship of paired images and use template matching to reduce MM of the dataset. Based on the CycleGAN framework, the synthesized loss and the generative reconstructed loss are added to the adversarial loss and the cycle-consistency loss to form a new objective function. And a claw-connected network is invoked to replace the U-net network structure of the generator for more feature preservation. The model is judged from subjective evaluation and objective evaluation based on image quality metrics such as PSNR (Peak Signal to Noise Ratio), SSIM (Structural Similarity), FID (Frechet inception distance), and face recognition accuracy. We divided the open datasets into bright light and dark light to research the effect of illumination. The experiments show that the proposed method has the lowest FID and the highest face recognition accuracy compared to the state-of-the-art methods. The proposed ClawGAN retains the structural features of thermal images while not only enhancing the quality of images but also effectively improving the observability of image translation results in both bright and dark light. The code is available at https://github.com/Luoyi3819/ClawGAN.

Automatic detection of mass in mammograms is a big challenge and plays a crucial role in assisting radiologists for accurate diagnosis. Applying excellent deep learning based detection networks to medical image is a direction. In this paper, an anchor-free YOLOv3 network is proposed for mass detection in mammogram. The main improvements are as follows. Anchor-based (AB) detection scheme may involve too many negative boxes, which may aggravate the unbalance between positive boxes and negative boxes. Aiming at this problem, a novel anchor-free (AF) detection scheme is developed for box regression. To alleviate the problem caused by using mean square error (MSE) as box regression loss, generalized intersection over union (GIoU) loss is adopted to train bounding box regression. Focal loss is adopted as the objectness prediction loss, which can prevent the network degradation caused by the vast number of easy negatives. Moreover, a new feature aggregation manner by summing is designed and employed in the up to down pathway. Comparative experiments are performed on two databases (publicly available dataset INbreast and private dataset TXMD). For INbreast dataset, the proposed method achieved 0.95 true positive rate (TPR) with 1.7 false positives per image (FPs/I). For TXMD dataset, the proposed method achieved 0.94 TPR with 5.75 FPs/I. Experimental results demonstrate that the proposed method outperforms YOLOv3 network and gets a promising result compared with existing mass detection methods.

The real-time pricing (RTP) plays an important role in demand side management of smart grid. In this paper, a distributed RTP strategy which takes the interests of both supply and demand sides into consideration is studied. A holistic model focusing on the interactions between users and the power supplier is proposed in the framework of Markov decision process (MDP). The MDP presentation of smart appliances’ operational processes well embodies their characteristics and the energy correlation of adjacent time slots. Different from existing ones, a novel distributed online algorithm based on a reinforcement learning approach is proposed to solve the MDP model without acquisition of the transition probabilities. Through information exchange between users and the power supplier, the real-time electricity price is decided adaptively, meanwhile, the optimal strategy of power supply and consumption is obtained. The specific information of the utility function for each user does not need to be disclosed to the supplier and other users. Simulation results show that the proposed model and algorithm balance energy supply and demand well, and have a good performance in peak shaving and valley filling.

Every larger organisation must establish a set of normative documents to control its processes and describe solutions to common problems. These documents are usually formally written and hard to read. This leads to the necessity of different customer services. Nowadays, a lot of companies are developing chatbots to automate first-line customer support. If a company does not have a large question-answer dataset to build a chatbot, the answers can be automatically answered directly from the documents. However, we found that the automatic answering usually does not work well on the normative documents. In this paper, we describe a novel method for preprocessing of normative documents in order to use them for such automatic question answering. Our method efficiently exploits the strict document structure that is typical for normative documents. We increased the recall from 35% to 84% (for paragraph-size answers) on selected normative documents from university and bank domains.

People use knowledge on several cognitive tasks such as when they recognize objects, rank entities such as the alternatives in multi-criteria decision making, or for classification tasks of decision making / expert / intelligent systems. When people have sufficient relevant knowledge, they can make well-distinctive assessments among entities. Otherwise, they may exhibit some uncertainty. This paper establishes two differential equations, of which one is for the interaction between the knowledge level and the uncertainty level, and the other is for the interaction between the ignorance level and the uncertainty level. By solving these two differential equations under certain boundary conditions, one can derive that the proposed knowledge level indicator is equivalent to Wierman's knowledge granularity measure up to a constant (exactly, ln 2 ). Moreover, the knowledge level indicator and the ignorance level indicator are found to be in a complementary relationship with each other. That is, more knowledge implies less ignorance, and vice-versa. The results of this study bridge a critical gap that exists in the understanding of the concepts of knowledge and ignorance.

Spherical fuzzy sets (SFSs) have recently become more popular in various fields. It was proposed as a generalization of picture fuzzy sets and Pythagorean fuzzy sets in order to deal with uncertainty and fuzziness information. The similarity measure is one of the beneficial tools to determine the degree of similarity between objects. It has many crucial application areas such as decision making, data mining, medical diagnosis, and pattern recognition. In the short time since their first appearance, some different distance and similarity measures of SFSs have been proposed, but they are limited through the literature. In this study, some novel distances and similarity measures of spherical fuzzy sets are presented. Then, we propose the Minkowski, Minkowski-Hausdorff, Weighted Minkowski and Weighted Minkowski-Hausdorff distances for SFSs. In addition, trigonometric and f-similarity measures are developed based on the proposed distances in this paper. The newly defined similarity measures are applied to medical diagnosis problem for COVID-19 virus and results are discussed. A comparative study of new similarity measures was established and some advantages of the proposed work are discussed.

Today’s organizations are witnessing a growing complexity in making interconnected decisions. Where individuals have a wider range of decisions to influence, the consequence of decisions far more propagate across the system, and the business environment continually influences the status of the system. Predicting the cascading effects of decisions in such situations would be very problematic yet can have several implications for managers and executives to think beyond organizational silos and make local decision with a bigger picture of emergent consequences in mind. A prominent challenge within this realm is the ever-increasing complexity of decision propagations, especially when incorporating the role and influence of people involved in decision-making. This paper tackles this challenge from an engineering change perspective, with the focus on computing the compound risk of decisions when their consequences concurrently propagate across the system. We introduce an interactive tool called Virtual Net Propagator, which incorporates organizational dynamics into decision analysis, with the aim to identify change opportunities and effective set of interventions. Illustrated by a field engineering case study, it is demonstrated that the proposed tool can provide detailed knowledge on how decisions are interconnected and how systemic (cascading) effects of a decision propagate through causal pathways, so highlighting key influencers along with role and influence of interfacing (hidden) players.

The condition of an inaccessible gear in an operating machine can be monitored using the vibration signal of the machine measured at some convenient location and further processed to unravel the significance of these signals. This paper deals with the effectiveness of wavelet-based features for fault diagnosis using support vector machines (SVM) and proximal support vector machines (PSVM). The statistical feature vectors from Morlet wavelet coefficients are classified using J48 algorithm and the predominant features were fed as input for training and testing SVM and PSVM and their relative efficiency in classifying the faults in the bevel gear box was compared.

This paper addresses the problem of classification when target data are subject to feature uncertainties. A robust approach based on Sobol sensitivity analysis is proposed to improve the robustness of support vector machine (SVM) models. SVM is a supervised machine learning method for pattern recognition whose performance depends on the definition of its hyperparameters and the quality of data. The proposed approach analyzes the impact of the uncertainties on the predictive performance of SVM based on Sobol’ sensitivity analysis. Afterwards, a new parameter is introduced to improve the robustness of SVM to the impact of uncertainties. The efficiency of this approach is evaluated by applying it to six real-world datasets. The results are then discussed and analyzed.

This paper aims to solve the large-scale heterogeneous container loading problem (HCLP), which is an extensive form of the multiple container loading problem, in a limited time. The target is to choose a set of containers of different sizes to accommodate all products and minimize the wasted space rate. Although the heterogeneous container selection problem is a general problem in the logistics industry, few related studies have been conducted. This study also considers some practical constraints, such as weight limits and suspension constraints. A hybrid adaptive large neighborhood search (HALNS) algorithm, which includes a set of original destroy-repair operators, especially for heterogeneous container selection problems, and integrates a heuristic packing algorithm, is proposed to solve the problem in an acceptable time. To verify the efficiency of the proposed algorithm, computational experiments on real-world instances from a multinational logistics company are performed, and the results are compared with those of other existing algorithms. The results indicate that the proposed algorithm outperforms other algorithms for the HCLP.

Predictive state representation~(PSR) uses a vector of action-observation sequence to represent the system dynamics and subsequently predicts the probability of future events. It is a concise knowledge representation that is well studied in a single-agent planning problem domain. To the best of our knowledge, there is no existing work on using PSR to solve multi-agent planning problems. Learning a multi-agent PSR model is quite difficult especially with the increasing number of agents, not to mention the complexity of a problem domain. In this paper, we resort to tensor techniques to tackle the challenging task of multi-agent PSR model development problems. By first focusing on a two-agent setting, we construct the system dynamics matrix as a high order tensor for a PSR model, learn the prediction parameters and deduce state vectors directly through two different tensor decomposition methods respectively, and derive the transition parameters via linear regression. Subsequently, we generalize the PSR learning approaches in a multi-agent setting. Experimental results show that our methods can effectively solve multi-agent PSR modelling problems in multiple problem domains., 20 pages, 16 figures

Face sketch synthesis plays a crucial role in face recognition for law enforcement applications. However, the current face sketch synthesis approaches generate sketches from photos based on a model trained by a certain database that is usually collected from individuals of the same ethnicity, and therefore such sketches merely inherit distinct facial distributions (shape and texture) of this database. This also makes such models inapplicable for real-world applications which mainly include multiple photo variations such as pose, lighting, skin color, and ethnic origin. In this paper, a unified face sketch synthesis model considering ethnicity issue as well as photo variations is proposed. A new deep learning scheme is designed to handle the generic visual representation and global structure of the face. Towards the final objective, the recent success of deep residual blocks is exploited and incorporated into a plain feedforward network, termed as DResNet, to learn a regression model for face sketch synthesis. A heterogeneous database containing photos with lighting, ethnicity, hair and skin variations is utilized for training the DResNet model. Extensive subjective and objective evaluations showed the superiority of the proposed DResNet method on state-of-the-art face sketch synthesis methods. Experimental results also demonstrated that the proposed DResNet method can be generalized to face sketch synthesis for real-world applications.

Class-imbalance learning is an important research area and draws continued attention in various real-world applications for many years. Undersampling is a key method of class-imbalance learning in order to obtain a balanced class distribution, but it may discard potentially crucial samples and may be influenced by outliers or noises in imbalanced data. Multiview learning methods have shown that models trained on different views can help each other to improve their performances and robustness, but the existing imbalance learning approaches rely only on single-view samples. In this paper, we propose a m ultiview feature imbalance sampling method via m eta s elf- p aced l earning (M2SPL) to effectively choose high-quality samples and separate adjacent features to improve the robustness of the trained model. There are two advantages of our proposed method: (1) An adaptive reweight generation process acts as a pivotal part in our M 2 SPL. The adaptive density-based reweight samples learning mechanism considers noisy and intractable samples to improve the robustness of model. (2) The multiview feature learning can avoid the large value of the loss function to learn a robust model from original data, and can enhance the discrimination capability of the model. Comparison with the existing sampling approaches shows that our proposed M2SPL approach significantly improves the classification performance, with increases in the F1-score and G-mean of 15.4% and 12.5%, respectively, on average. Finally, our experimental results pass the Friedman and Holm tests, indicating that our model has a significant improvement over existing methods.

Machine learning method has been widely applied in industrial fault diagnosis, especially the deep learning method. In the field of industrial fault diagnosis, deep learning is mostly used to extract features of vibration signals to achieve end-to-end fault diagnosis systems. Due to the complexity and variety of actual industrial datasets, some deep learning models are designed to be complicated. However, designing neural network architectures requires rich professional knowledge, experience, and a large number of experiments, increasing the difficulty of developing deep learning models. Fortunately, Neural Architecture Search (NAS), a branch of Automated Machine Learning (AutoML), is developing rapidly. Given a search space, NAS can search for networks that perform better than manually designed. In this paper, a one-shot NAS method for fault diagnosis is proposed. The one-shot model is a supernet that contains all candidate networks in a given search space. The supernet is trained to evaluate the actual performance of candidate networks by measuring the difference between its output probability and the true labels. According to the prediction of supernet, the networks with excellent performance can be searched, using some common search methods such as random search or evolutionary algorithm. Finally, the searched network is trained by reusing the weights of the supernet. To evaluate the proposed method, two search spaces are designed, ResNet and Inception search spaces, to search on PHM 2009 Data Challenge gearbox dataset. The state-of-the-art results are obtained, and accuracies of searched ResNet-A and Inception-A are 84.11% and 83.81%, which are 3.29% and 10.88% higher than Reinforcement Learning based NAS.

In recent years, there has been a surge of interest in applying machine learning to graphs and networks that already exist in the world around us. The approach has been successfully used for domains as diverse as traffic management, e-commerce recommendation and public opinion monitoring. A critical aspect of representation learning for applied machine learning is feature engineering. Deep learning-based research in representation learning has developed methods for automatically learning a large number of potentially correlated features from original networks. However, most of these methods cannot be applied to heterogeneous networks, which are true expressions of the real-world. This is because they do not adequately capture the structure and semantics of different types of nodes in heterogeneous networks and the links between them. They also struggle to represent higher-order heterogeneous patterns of connection. This paper proposes a generalized motif-based higher-order representation learning method, MBRep, that learns triangle motif embedding in a network, on the basis of which it can obtain the embedding and representation of nodes in a heterogeneous network. Statistically, significant motif structures are extracted from the original heterogeneous network and combined with the heterogeneity of the nodes. A weight-biased random walk is then applied to the motif level higher-order network, using a SkipGram model to embed the motifs. After this, the embedding of the original network nodes is calculated using weighted averages and feature alignment. This can then be used for link prediction. We confirmed the effectiveness of MBRep by comparing its AUC and MRR performance with other state-of-the-art methods on three real-world datasets. Its adaptability was also validated by conducting a cold-start test.

The network coverage problem for an area with a randomly deployed wireless sensor network (WSN) in the presence of obstacles can be alleviated through mobile sensor nodes. However, the requirement for the dissipated energy in mobility and sensing has increased. Therefore, energy efficient coverage is a significant issue and has been near-optimally solved by heuristic techniques. Focusing on this issue, this paper introduces an improved competitive swarm optimizer to maximize the coverage area and minimize the network energy simultaneously. The proposed method incorporates the virtual force algorithm (VFA) and the Voronoi diagram (VD) to improve the network performance during the optimization process. The VFA is combined with a boundary mechanism to control the locations of sensors, while the VD is utilized to extract the network information for the decoding process. The superior performance of the model is verified by intensive evaluations against state-of-the-art techniques in terms of the coverage ratio and network energy consumption.

Recent deep learning based facial expression recognition (FER) methods are mostly driven by the availability of large amount of training data. However, availability of such data is not always possible for FER in the wild where the infeasibility of obtaining sufficient training samples for each emotion category. Therefore, in this paper, we introduce the few-shot learning to construct a deep learning model named Convolutional Relation Network (CRN) for FER in the wild, which is learnt by exploiting a feature similarity comparison among the sufficient samples of the emotion categories to identify new classes with few samples. Specifically, our method learns a metric space in which classification can be performed by computing distances to capitalize on powerful discriminative ability of deep expression features to generalize the predictive power of the network. To achieve this, the features are constrained to maximize the distance between the features of different classes and discover the commonality of the same classes. Extensive experiments on three challenging in-the-wild datasets demonstrate that the proposed model significantly outperforms state-of-the-art methods.

Automated visual evaluation (AVE) is an emerging method to detect and diagnose cervical precancerous lesions by imaging and analysis via a deep learning classifier. Challenges in AVE development come from not only the limited data available, but also a proper design of the learning protocol. The most analyzed dataset (PEG) is traced to a clinical trial at a single site, where all the images were captured in a well controlled environment. Recently, cervical images have been captured by a light-weight mobile solution where the screening images were collected from a wider user pool at many sites. This paper introduces a new data resource (EVA dataset), collected by providers using a mobile colposcope during their routine practice. Compared to PEG, EVA images contain higher levels of data variations and exhibits a different distribution over multiple image attributes including image sharpness, brightness and colorfulness. In order to evaluate the practical value of EVA dataset for cervical high-grade squamous intraepithelial lesions (SIL) diagnosis, we further present an analysis of how a deep learning based framework can be used with both datasets by evaluating three key technical components: (1) region-of-interest (ROI) detection, (2) data augmentation and (3) pre-trained deep learning model selection. Our results indicate that ROI detection and shallow deep learning models usually help the detection on both datasets. While most data augmentations are effective on the EVA dataset, the improvement is less pronounced on PEG. Overall, using a deep-learning based framework looks promising for high-grade SIL diagnosis but there is still large room for improvements, especially on the EVA dataset. The noted differences indicate that the EVA dataset presents more practical challenges for high-grade SIL diagnosis and AVE classifier development.

The longest common consecutive subsequences (LCCS) play a vital role in revealing the biological relationships between DNA/RNA sequences especially the newly discovered ones such as COVID-19. FLAT is a Fragmented local aligner technique which is an accelerated version of the local pairwise sequence alignment algorithm based on meta-heuristic algorithms. The performance of FLAT needs to be enhanced since the huge length of biological sequences leads to trapping in local optima. This paper introduces a modified version of FLAT based on improving the performance of the BA algorithm by integration with particle swarm optimization (PSO) algorithm based on a novel infection mechanism. The proposed algorithm, named BPINF, depends on finding the best-explored solution using BA operators which can infect the agents during the exploitation phase using PSO operators to move toward it instead of moving toward the best-exploited solution. Hence, moving the solutions toward the two best solutions increase the diversity of generated solutions and avoids trapping in local optima. The infection can be propagated through the agents where each infected agent can transfer the infection to other non-infected agents which enhances the diversification of generated solutions. FLAT using the proposed technique (BPINF) was validated to detect LCCS between a set of real biological sequences with huge lengths besides COVID-19 and other well-known viruses. The performance of BPINF was compared to the enhanced versions of BA in the literature and the relevant studies of FLAT. It has a preponderance to find the LCCS with the highest percentage (88%) which is better than other state-of-the-art methods.

Secure indexing, ranking, searching, and retrieval in encrypted domain without compromising privacy is a challenging task. This is even more difficult in the case of images. In this paper, we propose a faster, more efficient, and highly scalable secure content-based image retrieval scheme using clustering for cloud. In the proposed scheme, the image owner extracts MPEG-7 visual descriptors from all the images and clusters them for indexing. These image features and cluster centers encrypted using Asymmetric scalar-product preserving encryption are offloaded to the cloud along with encrypted images to ensure secure retrieval of images. We also propose a copy-deterrence mechanism to identify untrustworthy query users. The proposed scheme is highly scalable, up to seven times faster in searching and up to 23 times faster in indexing with better retrieval precision compared with the fastest state-of-the-art scheme. The proposed copy deterrence mechanism ensures better speed and perceptual quality of watermarked images with better or comparable extraction accuracy when compromised.

The development of social media has caused a boom in information sharing, but it also provides an ideal platform for publishing and spreading rumors. On social media platforms, there are a lot of comments, which contain the user’s most direct views and reactions to the post. They can be utilized as clues to detect rumors. Recently, some methods are proposed to detect rumors through post and comments which usually focus on content information. However, except for this, other information such as reply structure, mutual selection information between post and comments, topic drift within comments also can help detect rumors. In this paper, we propose a novel model, PostCom2DR, for rumor detection. In PostCom2DR, a reply graph between post and comments is first created. Then a bilevel GCN and self-attention mechanism are built to learn the representation of comments. Secondly, a post-comment co-attention mechanism is introduced to selectively fuse information, and this helps the model focus on more relevant information. Through the above modules, we can get the global representation of post and comments. In addition, a 1D CNN is built to capture the local topic drift on time series inside the comments. Finally, we concatenate the global representation and local representation for rumor detection. Extensive experiments conducted on Chinese and English datasets show that PostCom2DR significantly outperforms other state-of-the-art models on both rumor detection and early detection.

The low data quality of a cooperative information system (CIS) causes cooperative organizations to recollect data frequently rather than use the data already in the system. In this paper, a method based on source control is proposed to reduce the data quality problem of CIS. It emphasizes improving the data quality of CIS by controlling the data quality of subsystems rather than by screening data. Further, we develop two optimization models for the scenes of independent and negotiated decisions. Compared with the independent decision model, the negotiation decision model considers the impact of the game behavior between the manager of the CIS and the manager of each subsystem on the data quality optimization. The results of the models provide an optimal strategy for allocating capital to minimize the data quality problem of a CIS. To solve the models, we propose an improved artificial bee colony algorithm that enhances the exploration ability by modifying the behaviors of bees. A CIS of a government is used to illustrate the application process of the models and the superiority of the algorithm. The results show that, in all cases, the optimal strategy provided by the negotiation decision model is not worse than that offered by the independent decision model, and the result obtained by the proposed algorithm is better than other comparative algorithms.

With the huge expansion of user generated content on social networks, event detection has emerged as a major challenge and source of knowledge discovery. This knowledge is employed in different applications such as recommender systems, crisis management systems, and decision support systems. Dynamicity, overlapping, and evolutionary behavior are the most important issues in event detection. This paper proposes a novel evolutionary model for event detection to capture the dynamism and evolving behavior of events. The proposed method uses a matrix decomposition technique and a Dirichlet Process to detect events and handle their dynamicity. This model consists of two components, namely preliminary event detection and event evolvement tracking. The former component extracts preliminary events from the available data using the matrix decomposition method. Then, subsequent data is employed into a Non-Parametric Bayesian Network, namely Dirichlet Process Mixture Model to evolve the preliminary events. During the evolvement process, data may migrate between extracted events or new events may be discovered. The experimental results and comparisons with several recently developed approaches show the superiority of the proposed approach, and its ability to capture the evolutionary behavior of events over time.

Interpretable machine learning is trending as it aims to build a human-understandable decision process. There are two main types of machine learning systems: white-box and black-box models. White-box models are inherently interpretable but commonly suffer from under-fitting phenomena; on the other hand, black-box models perform quite well in a wide range of application domain problems, but their reasoning behind a decision is hard or even impossible to understand. In the soft-computing area, fuzzy inference systems are rule-based systems that use fuzzy reasoning, bringing human perception modeling and computing with word capability. These rule-based systems are designed either manually or automatically but are commonly optimized to fit better some phenomena’ data (in a supervised learning task). After the optimization process, the initial semantic meaning of fuzzy sets is modified (slightly, in the best cases), creating a gray-box model. The principal objective of the proposed methodology in this paper is to extract a high-quality rule in terms of comprehensibility, accuracy and fidelity. This is accomplished by using a fuzzy linguistic interpretable model from an optimized neuro-fuzzy model, considering the initial knowledge context with which it was built. A grammar-guided genetic algorithm is used as the optimization process to find the interpretable description of the model. A collection of 16 datasets for classification tasks were used to evaluate our proposal, obtaining an f1-score of 0.814 with 0.026 standard deviation in the optimized model; the obtained fidelity, in terms of similarity from the interpretable model to the optimized one, was 0.93 of mean with 0.018 standard deviation. Obtained results show that neuro-fuzzy systems could play an important role in interpretable machine learning, providing natural language explanations from previous knowledge.