Search

Academic research is crucial to the development of science and technology and is an important factor that affects national strength. When writing an academic research paper, a rhetorical structure is typically used to present the paper’s ideas, but this task is quite difficult for junior researchers. To solve this problem, some studies have adopted text mining to assist with the writing, but the existing methods still require human intervention to generate sentences. Recently, due to the increasing maturity of deep learning technology and the ability to address the problem of automatic text generation, progress has been made in this area. The highly complex deep learning operations can correctly generate sequences and find correlations between sequences. When a user provides a few keywords and key sentences, the proposed algorithm can generate an introduction section for the user. The results show that the generated introduction is more coherent, clearer, and more fluent than existing summarization methods. In addition, the method proposed in this study improves the accuracy compared with traditional text extraction methods. The manuscript produced by this study has been evaluated to show that the study can produce a comprehensive introduction compared with previous studies.

This paper provides a corrected formulation to the mixed integer programming model proposed by Aydogan et al. (2012) 1. They proposed a genetic algorithm to learn fuzzy rules for a fuzzy rule-based classification system and developed a Mixed Integer Programming model (MIP) to prune the generated rules by selecting the best set of rules to maximize predictive accuracy. However, their proposed MIP formulation contains errors, which are described in this technical note. We develop corrections and improvements to the original formulation and test it with non-parametric statistical tests on the same data sets used to evaluate the original model. The statistical analysis shows that the results of the correction formulation are significantly different from the original model.

Related works for applying keystroke dynamics (KD) on free text identification indicated that applying KD can improve the accuracy of personal authentication on free text. As the result, this paper proposes a new biometrics, i.e., the keystroke clusters map (KC-Map), by clustering users’ keystrokes in order to effectively enhance the accuracy of personal authentication in free text. Since KC-Map is conducted via clustering, it is not suitable for traditional classifiers. In order to tackle this problem, the paper further proposes a keystroke clusters map similarity classifier (KCMS classifier). Experimental results positively show that the proposed KC-Map and KCMS classifier can efficiently improve the accuracy of personal authentication on free text with up to 1.27 times. In addition, one of the huge disadvantages on the current approaches in free text identification is that users are generally required to be trained for several months. Longer training time makes free text identification more impractical. Another motivation of this paper is to explore whether it is possible to shorten the training time into an acceptable range. Experimental results show that, to achieve relatively fair identification accuracy, users only need to carry out about 20 min for training.

This paper considers the lot scheduling problem in the flexible flow shop with limited intermediate buffers to minimize total cost which includes the inventory holding and setup costs. The single available mathematical model by Akrami et al. (2006) for this problem suffers from not only being non-linear but also high size-complexity. In this paper, two new mixed integer linear programming models are developed for the problem. Moreover, a fruit fly optimization algorithm is developed to effectively solve the large problems. For model’s evaluation, this paper experimentally compares the proposed models with the available model. Moreover, the proposed algorithm is also evaluated by comparing with two well-known algorithms (tabu search and genetic algorithm) in the literature and adaption of three recent algorithms for the flexible flow shop problem. All the results and analyses show the high performance of the proposed mathematical models as well as fruit fly optimization algorithm.

In recent years, smog weather in China has continued to worsen. Aiming at the problem of smog pollution prediction, a new grey extended prediction model, which is based on the GM (grey model) (1,1) model, is proposed in this paper. When the input function takes different forms, the grey extended model can degenerate into various existing grey prediction models. At the same time, the iterative reweighted least squares method (IRLS) is introduced to improve the models. In addition, to study the application effect of the grey extended prediction model GM(1,1,u(t)) and the iterative reweighted least squares method, based on the data of the air quality index (AQI) from December 2013 to February 2018, in Shanghai, four different forms of the GM(1,1,u(t)) model are established in this paper. The ordinary least square method (OLS) and the iterative reweighted least squares method are used to estimate the parameters, and the simulation and prediction results of the models are compared with those of the univariate linear regression model, the autoregressive integrated moving average (ARIMA) model and the feedforward neural network model. The result shows that the GM(1,1,u(t)) model (where u(t) = 1) under the IRLS method has the best prediction effect, which is not only suitable for predicting the normal values of AQI, but also suitable for predicting the outliers of AQI Therefore, it is proven that the GM(1,1,u(t)) model under iterative reweighted least squares is superior in simulation and prediction.

This paper proposes an adaptive S transform (AST) to extract the feature vectors of voltage sags. With the effective window width matches the Fourier spectrum of sag signals, the standard deviation σ of Gaussian window may be determined as well. The narrowest and the widest window width of AST are obtained without additional parameters and iterative computing. Then, the optimal frequency resolution and time resolution are got respectively. Compared with ST, AST provides better time–frequency resolution to extract more precise feature vectors of eight types of voltage sags. Based on the time–frequencyrepresentation of AST, five disturbance features are extracted to construct the feature vector in this paper. In addition, four machine learning classifiers and two fuzzy clustering classifiers are used to analyze the validity and redundancy of these features. Through analyzing the classification accuracies and time costs of these classifiers with different training sets and different level of noise, it can be concluded that the machine learning classifiers perform better in classification accuracy and stability than fuzzy clustering classifiers.

Differential evolution is primarily designed and used to solve continuous optimization problems. Therefore, it has not been widely considered as applicable for real-world problems that are characterized by permutation-based combinatorial domains. Many algorithms for solving discrete problems using differential evolution have been proposed, some of which have achieved promising results. However, to enhance their performance, they require improvements in many aspects, such as their convergence speeds, computational times and capabilities to solve large discrete problems. In this paper, we present a new mapping method that may be used with differential evolution to solve combinatorial optimization problems. This paper focuses specifically on the mapping component and its effect on the performance of differential evolution. Our method maps continuous variables to discrete ones, while at the same time, it directs the discrete solutions produced towards optimality, by using the best solution in each generation as a guide. To judge its performance, its solutions for instances of well-known discrete problems, namely: 0/1 knapsack, traveling salesman and traveling thief problems, are compared with those obtained by 8 other state-of-the-art mapping techniques. To do this, all mapping techniques are used with the same differential evolution settings. The results demonstrated that our technique significantly outperforms the other mapping methods in terms of the average error from the best-known solution for the traveling salesman problems, and achieves promising results for both the 0/1 knapsack and the traveling thief problems.

Cloud detection algorithms have emerged to automate image data analysis because of its prime influential factor in remote sensing image quality. Cloud detection algorithm still needs domain-expert intervention and large number of training examples to ensure good performance whose acquirement becomes difficult due to unavailability of labeled data as well as the time and process heads involved. The paper puts forward multi-objective social spider optimization (MOSSO) based efficient clustering technique to detect clouds in the visible range. This paper explains the proposed MOSSO algorithm along-with the analysis carried on 14 benchmark two-objective test problems against MOEA/D, MODE, MOPSO and SPEA2 multi-objective algorithms. Further, the strengths and weaknesses of the proposed algorithm are analyzed and have been used for the implementation of an efficient clustering technique named as MOSSO-C. Optimal centroid matrix for clustering is attained in MOSSO-C through environmental selection whose performance evaluation has been done on six synthetic databases and are compared with above mentioned conventional multi-objective algorithms. The obtained results encourage the use of MOSSO-C technique to get labeled data for training process of neural network classifier. This approach efficiently classifies the cloudy pixels against various Earth’s surfaces (water, vegetation and land). The paper also discusses the performance evaluation of proposed technique on four Landsat 8 data which shows on an average 96.37% performance accuracy in detecting cloudy pixels.

Regression problems try estimating a continuous variable from a number of characteristics or predictors. Several proposals have been made for regression models based on the use of fuzzy rules; however, all these proposals make use of rule models in which the irrelevance of the input variables in relation to the variable to be approximated is not taken into account. Regression problems share with the ordinal classification the existence of an explicit relationship of order between the values of the variable to be predicted. In a recent paper, the authors have proposed an ordinal classification algorithm that takes into account the detection of the irrelevance of input variables. This algorithm extracts a set of fuzzy rules from an example set, using as the basic model a sequential covering strategy along with a genetic algorithm. In this paper, a proposal for a regression algorithm based on this ordinal classification algorithm is presented. The proposed model can be interpreted as a multiclassifier and multilevel system that learns at each stage using the knowledge gained in previous stages. Due to similarities between regression and ordinal problems as well as the use of a set of ordinal algorithms, an error interval can be returned with the regression output value. Experimental results show the good behavior of the proposal as well as the results of the error interval.

Demand response (DR) is the response of electricity consumers to time-varying tariffs or incentives awarded by the utility. Home energy management systems are systems whose role is to control the consumption of appliances under DR programs, in a way that electricity bill is minimised. While, most researchers have done optimal scheduling only for non-interruptible appliances, in this paper, the interruptible appliances such as electric water heaters are considered. In optimal scheduling of non-interruptible appliances, the problem is commonly formulated as an optimisation problem with integer decision variables. However, consideration of interruptible appliances leads to a binary optimisation problem which is more difficult than integer optimisation problems. Since, the basic version of binary particle swarm optimisation (PSO) does not perform well in solving binary engineering optimisation problems, in this paper a new binary particle swarm optimisation with quadratic transfer function, named as quadratic binary PSO (QBPSO) is proposed for scheduling shiftable appliances in smart homes. The proposed methodology is applied for optimal scheduling in a smart home with 10 appliances, where the number of decision variables is as high as 264. Optimal scheduling is done for both RTP and TOU tariffs both with and without consideration of consumers’ comfort. The achieved results indicate the drastic effect of optimal scheduling on the reduction of electricity bill, while consumers’ comfort is not much affected. The results testify that the proposed QBPSO outperforms basic binary PSO variant and 9 other binary PSO variants with different transfer functions.

Leaky integrator echo state network (Leaky-ESN), an improved echo state network, is suitable for learning very slow dynamic systems and replaying the learnt system at different speeds. To ensure the echo state property of Leaky-ESN, spectral radius of reservoir connection weight matrix needs to be less than leaking rate, which means that the output feedback of Leaky-ESN is actually ignored. Leaky-ESN without output feedback is equivalent to being a feedforward type neural network. Therefore, in this paper, we consider Leaky-ESN model with output feedback. Firstly, we give the modified sufficient conditions for echo state property, which are actually some inequality constraints about the control parameters (mainly leaking rate, spectral radius of reservoir connection weight matrix and the scaling of output feedback) and the maximum singular values of feedback and output connection weight matrices. Secondly, to reduce the influence of the initial values of the control parameters on Leaky-ESN, we use the barrier method to optimize the control parameters. The barrier method can convert the constrained optimization problem into the unconstrained optimization problem. Newton’s method and its improve method, for example, eigenvalue modification methods, are further used to solve the unconstrained optimization problem and then obtain the control parameters of Leaky-ESN. Finally, Mackey–Glass chaotic time series prediction is selected to validate the method proposed in this paper. Simulation results show that the method proposed in this paper can make Leaky-ESN model more stable and higher approximation precision.

Wellbore trajectory design is a determinant issue in drilling engineering. The criteria to evaluate a wellbore trajectory are summarized as the total trajectory length, the torque and the well profile energy in this paper. By minimizing the wellbore trajectory length, torque and profile energy simultaneously, it is most likely that a wellbore trajectory designed to arrive at the specific target can be drilled more safely, quickly and cheaply than other potential trajectories. However, these three objectives are often in conflict with each other and related in a highly nonlinear relationship. A multi-objective cellular particle swarm optimization (MOCPSO) with an adaptive neighborhood function is developed in this paper. Then, MOCPSO is applied with the three objective functions to gain a set of Pareto optimal solutions that are beneficial for a less risky and less costly wellbore trajectory design option. Besides, MOCPSO’s performance is compared with multi-objective PSO, multi-objective evolutionary algorithm based on decomposition (MOEA/D) and non-dominated sorting genetic algorithm-II (NSGA-II). Effect of the proposed neighborhood function is also investigated by making contrasts with the commonly used four neighborhood templates. Moreover, the radius parameter in the adaptive neighborhood function is analyzed to reveal its influence on the optimization performance. It can be inferred that MOCPSO is statistically superior to both multi-objective PSO, NSGA-II and MOEA/D at the 0.05 level of significance on the wellbore trajectory design problem. And the proposed adaptive neighborhood function performs either comparable or better as compared to the other commonly used neighborhood functions. According to the parameter analysis, it can be concluded that the MOCPSO approach with radius value of 1or 1.5 has a better statistical performance.

The linguistic computing model based on discrete fuzzy numbers has some good properties compared with other existing models and should be further studied, which has been proved by some researchers. However, the research of group consensus with this linguistic model is insufficient, given that group consensus is an important issue in group decision making. Therefore, this paper would concentrate on this subject. It includes two main issues: research on consensus measure and research on the method for improving group consensus in group decision making based on this linguistic computing model. For research on the consensus measure, this paper first studies on the aggregation method for discrete fuzzy numbers. Then, the index of measuring group consensus is determined. For research on improving the group consensus, considering the characteristics of discrete fuzzy numbers, we present an algorithm to improve group consensus. In addition, an illustrative example of a decision-making problem about investment is stated to show the whole solving process. It also illustrates the feasibility, rationality and validity of all the proposed methods. Finally, the comparisons between some proposals and existing studies are made, which helps point out the advantages of the proposed methods.

Hesitant fuzzy linguistic preference relations (HFLPRs) as an efficient and common tool to deal with decision-making problems have been widely used in real life. The consistency and consensus are the most two important topics for HFLPRs. In this paper, we develop a new efficient consistency–consensus framework for HFLPRs. Firstly, we transform HFLPRs into probabilistic linguistic preference relations, and develop a programming model to make the HFLPRs achieve the maximum consistency degree. Then, we propose some new rules for addition operation and weighted average operator to fuse individual preference information. After that, a maximum consensus model is developed which maximizes the consensus degree by adjusting the experts’ weights. Then, a numerical case of temporary placement selection after earthquake is presented, and some comparisons with the traditional methods are conducted to verify the rationality and efficiency of our framework. Finally, we end the paper with some conclusions and future research directions.

Interval type-2 fuzzy logic controllers (IT2-FLCs) have been attracting a lot of attention. However, challenges in designing IT2-FLCs still remain. One of the main challenges is to choose the appropriate FOU shape for interval type-2 fuzzy sets (IT2-FSs). This paper aims to analyse the differences in control performance between three IT2 fuzzy PI controllers (IT2-F-PICs) with different FOU shapes as antecedent sets, namely the triangular top wide IT2 fuzzy set, the triangular bottom wide IT2 fuzzy set and the trapezoidal (also called parallel) IT2 fuzzy set. First, the analytical structures of these IT2-FLCs are derived and the mathematical input–output equations are obtained. Three interesting differences between the analytical structures and input–output relationship of the IT2-F-PICs are then presented. From the differences in the analytical structures of the three IT2-F-PICs and numerical simulation results, it is demonstrated that IT2-F-PICs with trapezoidal (IT2-F-PI-P) and triangular bottom wide (IT2-F-PI-BW) antecedent sets with the potential to provide faster transient response and faster settling time than the IT2-F-PICs with triangular top wide (IT2-F-PI-TW). In addition, IT2-F-PI-P is better able to handle plant uncertainties and disturbances than IT2-F-PI-BW and IT2-F-PI-TW. The contribution of this paper is to provide insights into the performance differences between different FOU shaped controllers, which in turns allowing control designers to select the appropriate FOU shape in order to meet design requirements.

In this paper, we propose a new method of computing an approximate Nash equilibrium with additional features. Existing algorithms often fail to produce an exact solution for games involving more than 3 players. Similarly, existing algorithms do not permit additional constraints on the problem. The principle idea of this paper involves proposing a methodology for computing approximate solutions through evolutionary computation. To do so, we first provide formal definitions of these problems and their approximate versions. Following which, we present the details of our solution. One of the most important advantages of the proposed solution is flexibility, which provides solutions to problems related to Nash equilibrium extensions. The proposed idea is tested on several types of games that vary with difficulty and size. All test sets are generated based on the well-known Gamut program. Additional comparisons with classical algorithms are also performed. Results indicate that Differential Evolution is capable of obtaining satisfactory solutions to large random and covariant games. The results also demonstrate that there is a high probability that even large games, in which a set of strategies with a non-zero probability of being chosen are very small, have a solution. The computation time depends mainly on the problem size, and the original Nash equilibrium problem is unaffected by additional modifications.

Constrained optimization problems arise in numerous scientific and engineering applications, and many papers on the online solution of constrained optimization problems using projection neural networks have been published in the literature. The purpose of this paper is to provide a comprehensive review of the research on projection neural networks for solving various constrained optimizations as well as their applications. Since convergence and stability are important for projection neural networks, theoretical results of projection neural networks are reviewed in detail. In addition, various applications of projection neural networks, e.g., the motion generation of redundant robot manipulators, coordination control of multiple robots with limited communications, generation of winner-take-all strategy, model predictive control and WSN localizations, are discussed and compared. Concluding remarks and future directions of projection neural networks as well as their applications are provided.

Brain–Computer Interfaces (BCI) are systems that translate brain activity patterns into commands for an interactive application, and some of them recognize patterns generated by motor imagery. Currently, these systems present performances and methodologies that still are not practical enough for realistic applications. Therefore, this paper proposes two methodologies for multiple motor imagery classification. Both methodologies use features extracted by a variant of Discriminative Filter Bank Common Spatial Pattern (DFBCSP) presented in this paper. The frequency bands selection in this variant is carried out by a novel iterative algorithm that selects the frequency band that attains the highest classification accuracy for specific binary classification. For each binary combination of classes, a frequency band is selected. The resulting samples are then set into a matrix which feeds one or many Convolutional Neural Networks previously optimized by using a Bayesian optimization. The first methodology applies a Convolutional Neural Network (CNN) for the classification of all classes and the second is a modular network composed of four expert CNNs. In this modular network, each expert CNN performs a binary classification, and a fully connected network analyzes their results. To validate both approaches two datasets were used, the BCI competition IV dataset 2a and another presented in this paper recorded from eight subjects by using the OpenBCI device. The experimental results demonstrated an improvement in the classification accuracy over many classic intelligent recognition methods, without a high computation time in order that they can be implemented in an online application.

The importance, benefits, and impact of integration of decisions within supply chains have long been investigated by many researchers. Order acceptance and supplier selection are two of the most critical decisions for supply chain managers. Throughout the process of order acceptance, a manufacturer has to decide which orders to be accepted and processed and based on the accepted orders, the volume of required raw material is determined. On the other hand, a manufacturer aims to choose one or several suppliers among all possible choices to provide sufficient raw material for the accepted orders, subject to different criteria such as list price, transportation cost, etc. This paper addresses an integrated framework for profit maximization in an integrated supplier selection, order acceptance and scheduling problem in a single-machine environment with multiple customers. There is substantial literature on the problems of supplier selection and order acceptance; however, to the best of our knowledge, this paper is the first research that integrates these essential decisions in the form of a mathematical model to maximize the total profit. The problem is NP-hard in nature; therefore, solving to optimality is not practically possible for problems with medium and large size. For that purpose, we developed a Heuristic Algorithm (HA) to solve the problem above in a reasonable time, with proper accuracy. Results from this heuristic algorithm are compared with that of a commercial solver (GAMS) and the well-known Genetic Algorithm (GA) and Variable Neighborhood Search (VNS). Computational experiments demonstrate that the developed heuristic algorithm is more efficient in comparison with other tested methods.

In this paper, a new nonlinear heuristic filter based on the hybridization of an extended Kalman filter and an ant colony estimator is proposed to estimate the states of a nonlinear system. In this filter, a group of virtual ants searches the state space stochastically and dynamically to find and track the best state estimation while the position of each ant is updated at the measurement time using the extended Kalman filter. The performance of the proposed filter is compared with well-known heuristic filters using a nonlinear benchmark problem. The statistical results show that this algorithm is able to provide promising and competitive results. Then, the new filter is tested on a nonlinear engineering problem with more than one state. The problem is to estimate simultaneously the states of an unmanned aerial vehicle as well as the wind disturbances, applied to the system. In this case, a processor-in-the-loop experiment is also performed to verify the implementation capability of the proposed approach. This paper also investigates the real-time implementation capability of the proposed filter in the attitude estimation of a three degrees of freedom experimental setup of a quadrotor to further investigate its effectiveness in practice.

Hybrid optimization algorithms which combine Evolutionary Algorithms (EAs) and Gradient-Based (GB) methods o refine the most promising solutions, are valuable tools for use in engineering optimization. Several hybrid methods can be found in the literature; differences among them are associated with the criteria used to select individuals for refinement through the GB method and the feedback the EA gets from the latter. GB methods require the gradient of the objective functions with respect to the design variables. By employing the adjoint method in problems governed by partial differential equations, the cost of computing the gradient becomes independent of the number of design variables. For multi-objective optimization problems this paper is exclusively dealing with, the availability of the gradients of all objective functions is not enough. Hybrid algorithms require the computation of descent directions in the objective space capable of improving the current front of non-dominated solutions. Using the sum of weighted objectives as the new objective function is ineffective. In this paper, a method which refrains from using arbitrarily defined weights is proposed. The method is driven by data obtained from the Principal Component Analysis (PCA) of the objective function values of the elite individuals at each generation of the EA. The PCA, with computational cost that of the solution of an eigenproblem, identifies the direction in the objective space along which the current front of non-dominated solutions should be improved. This along with the gradients computed by the adjoint method are used by the GB method to refine selected individuals. The efficiency of the proposed hybrid algorithm is further improved by employing online trained surrogate models or metamodels and Kernel PCA within the EA-based search. The proposed method is demonstrated in aerodynamic shape optimization problems, using in-house Computational Fluid Dynamics software and its adjoint.

Missing values are an unavoidable issue in many real-world datasets. One of the most popular approaches to classification with incomplete data is to use imputation to replace missing values with plausible values. However, powerful imputation methods are too computationally intensive when applying a classifier to a new unknown instance. This paper proposes new approaches to integrating imputation, clustering and feature selection for classification with incomplete data in order to improve efficiency without loss of accuracy. Clustering is used to reduce the number of instances used by the imputation. Feature selection is used to remove redundant and irrelevant features of training data which greatly reduces the cost of imputation. The paper also investigates the ability of Differential Evolution (DE) to search feature subsets with incomplete data. Results show that the integration of imputation, clustering and feature selection not only improves classification accuracy, but also dramatically reduces the computation time required to estimate missing values when classifying new instances.

Credit scoring models have offered benefits to lenders and borrowers for many years. However, in practice these models are normally built on a sample of accepted applicants and fail to consider the remaining rejected applicants. This may cause a sample bias which is an important statistical issue, especially in the online lending situation where a large proportion of requests are rejected. Reject inference is a method for inferring how rejected applicants would have behaved if they had been granted and incorporating this information in rebuilding a more accurate credit scoring system. Due to the good performances of SVM models in this area, this paper proposes a new approach based on the state-of-the-art kernel-free fuzzy quadratic surface SVM model. It is worth pointing out that our method not only performs very well in classification as some latest works, but also handles some big issues in the classical SVM models, such as searching proper kernel functions and solving complex models. Besides, this paper is the first one to eliminate the bad effect of outliers in credit scoring. Moreover, we use two real-world loan data sets to compare our method with some benchmark methods. Particularly, one of the data set is very valuable for the study of reject inference, because the outcomes of rejected applicants are partially known. Finally, the numerical results strongly demonstrate the superiority of the proposed method in applicability, accuracy and efficiency.

Unsupervised classification is a crucial step in remote sensing hyperspectral image analysis where producing labeled data is a laborious task. Spectral Clustering is an appealing graph-partitioning technique with outstanding performance on data with non-linear dependencies. However, Spectral Clustering is restricted to small-scale data and neither has been effectively applied to hyperspectral image analysis. In this paper, the unsupervised classification of hyperspectral images is addressed through a sequential spectral clustering that can be extended to the large-scale hyperspectral image. To this end, this paper utilizes a bipartite graph representation along with a sequential singular value decomposition and mini-batch K-means for unsupervised classification of hyperspectral imagery. We evaluate the proposed algorithm with several benchmark hyperspectral datasets including Botswana, Salinas, Indian Pines, Pavia Center Scene and Pavia University Scene. The experimental results show significant improvements made by the proposed algorithm compared to the state-of-art clustering algorithms.

The majority of the studies on credit risk assessment models for financial institutions during recent years focus on the improvement of imbalanced data or on the enhancement of classification accuracy with multistage modeling. Whilst multistage modeling and data pre-processing can boost accuracy somewhat, the heterogeneous nature of data may affects the classification accuracy of classifiers. This paper intends to use the classifier, eXtreme gradient boosting tree (XGBoost), to construct a credit risk assessment model for financial institutions. Cluster-based under-sampling is deployed to process imbalanced data. Finally, the area under the receiver operative curve and the accuracy of classifications are the assessment indicators, in the comparison with other frequently used single-stage classifiers such as logistic regression, self-organizing algorithms and support vector machine. The results indicate that the XGBoost classifier used by this paper achieve better results than the other three and can serve as a superior tool for the development of credit risk models for financial institutions.

Background: The ongoing fight with Novel Corona Virus, getting quick treatment, and rapid diagnosis reports have become an act of high priority. With millions getting infected daily and a fatality rate of 2%, we made it our motive to contribute a little to solve this real-world problem by accomplishing a significant and substantial method for diagnosing COVID-19 patients. Aim: The Exponential growth of COVID-19 cases worldwide has severely affected the health care system of highly populated countries due to proportionally a smaller number of medical practitioners, testing kits, and other resources, thus becoming essential to identify the infected people. Catering to the above problems, the purpose of this paper is to formulate an accurate, efficient, and time-saving method for detecting positive corona patients. Method: In this paper, an Ensemble Deep Convolution Neural Network model “CoVNet-19” is being proposed that can unveil important diagnostic characteristics to find COVID-19 infected patients using X-ray images chest and help radiologists and medical experts to fight this pandemic. Results: The experimental results clearly show that the overall classification accuracy obtained with the proposed approach for three-class classification among COVID-19, Pneumonia, and Normal is 98.28%, along with an average precision and Recall of 98.33% and 98.33%, respectively. Besides this, for binary classification between Non-COVID and COVID Chest X-ray images, an overall accuracy of 99.71% was obtained. Conclusion: Having a high diagnostic accuracy, our proposed ensemble Deep Learning classification model can be a productive and substantial contribution to detecting COVID-19 infected patients.

COVID-19 originally known as Corona VIrus Disease of 2019, has been declared as a pandemic by World Health Organization (WHO) on 11th March 2020. Unprecedented pressures have mounted on each country to make compelling requisites for controlling the population by assessing the cases and properly utilizing available resources. The rapid number of exponential cases globally has become the apprehension of panic, fear and anxiety among people. The mental and physical health of the global population is found to be directly proportional to this pandemic disease. The current situation has reported more than twenty four million people being tested positive worldwide as of 27th August, 2020 Therefore, it’s the need of the hour to implement different measures to safeguard the countries by demystifying the pertinent facts and information. This paper aims to bring out the fact that tweets containing all handles related to COVID-19 and WHO have been unsuccessful in guiding people around this pandemic outbreak appositely. This study analyses two types of tweets gathered during the pandemic times. In one case, around twenty three thousand most re-tweeted tweets within the time span from1st Jan 2019 to 23rd March 2020 have been analysed and observation says that the maximum number of the tweets portrays neutral or negative sentiments. On the other hand, a dataset containing 226668 tweets collected within the time span between December 2019 and May 2020 have been analysed which contrastingly show that there were a maximum number of positive and neutral tweets tweeted by netizens. The research demonstrates that though people have tweeted mostly positive regarding COVID-19, yet netizens were busy engrossed in re-tweeting the negative tweets and that no useful words could be found in WordCloud or computations using word frequency in tweets. The claims have been validated through a proposed model using deep learning classifiers with admissible accuracy up to 81%. Apart from these the authors have proposed the implementation of a Gaussian membership function based fuzzy rule base to correctly identify sentiments from tweets. The accuracy for the said model yields up to a permissible rate of 79%., Highlights • This paper aims to bring out the fact that related tweets have been unsuccessful in guiding people around the COVID-19 pandemic. • This study analyses two types of tweets gathered during the pandemic times. • The research demonstrates that no useful words could be a found in WordCloud or computations using word frequency in tweets. • The claims have been validated through a proposed model using deep learning classifiers with admissible accuracy up to 81%. • A Gaussian membership function based fuzzy rule base is also designed to correctly identify sentiments from tweets.

This paper employs a novel adaptive feature extraction techniques of electrocardiogram (ECG) signal for detection of cardiac arrhythmias using multiresolution discrete wavelet transform from ECG big data. In this paper, five types ECG arrhythmias including normal beats have been classified. The MIT-BIH database of 48 patient records is utilized for detection and analysis of cardiac arrhythmias. Proposed feature extraction utilizes Daubechies as wavelet function and extracts 21 feature points which include the QRS complex of ECG signal. The Multilayered Probabilistic Neural Network (MPNN) classifier is proposed as the best-suited classifier for the proposed feature. Total 1700 ECG betas were tested using MPNN classifier and compared with other three classifiers Back Propagation (BPNN), Multilayered Perceptron (MLP) and Support Vector Machine (SVM). The system efficiency and performance have been evaluated using seven types of evaluation criteria: precision (PR), F-Score, positive predictivity (PP), sensitivity (SE), classification error rate (CER) and specificity (SP). The overall system accuracy, using MPNN technique utilizing the proposed feature, obtained is 99.53% whereas using BPNN, MLP and SVM provide 97.94%, 98.53%, and 99%. The processing time using MPNN classifier is only 3 s which show that the proposed techniques not only very accurate and efficient but also very quick.

Optimization of complex real-world problems often involves multiple objectives to be considered simultaneously. These objectives are often non-commensurable and competing. For example, material handling is a vital element of industrial processes, which involves a variety of operations including the movement, storage and control of materials throughout the processes of manufacturing, distribution, and disposal while having to satisfy multiple objectives. Having an efficient and effective material handling system (MHS) is of great importance to various industries, such as manufacturing and logistics industries, for maintaining and facilitating a continuous flow of materials through the workplace and guaranteeing that required materials are available when needed. In this paper, a hybrid multi-objective optimization algorithm largely based on Artificial Immune Systems (AIS) is applied to simulation-based optimization of material handling system. This proposed algorithm hybridizes the AIS with the Genetic Algorithm (GA) by incorporating the crossover operator derived from the biological evolution. The reason behind such hybridization is to further enhance the diversity of the clone population and the convergence of the algorithm. In this paper, other than conducting numerical experiments to assess the performance of the proposed algorithm by using several benchmark problems, the proposed algorithm is also applied to optimize a multi-objective simulation-based problem on a material handling system in order to demonstrate the applicability of the proposed algorithm in real-life cases. The results show that for most cases the proposed algorithm outperforms the other benchmark algorithms especially in terms of solution diversity.

The aim of this paper is to develop a novel outranking method using a risk attitudinal assignment model under complex uncertainty based on Pythagorean fuzzy (PF) sets and interval-valued Pythagorean fuzzy (IVPF) sets. Additionally, this paper applies the proposed methodology to address financial decision-making problems to investigate its applicability and effectiveness in the real world. These developed methods and techniques can provide a new viewpoint for capturing the amount of information conveyed by PF and IVPF values and can fully take into account the decision maker’s attitude towards risk-taking in the face of high uncertainty. Considering the diverse attitudes toward risk neutrality, risk seeking, and risk aversion, this paper introduces novel measures of risk attitude-based score functions within the PF/IVPF environments and investigates their desirable and useful properties. These properties provide a means of establishing linear orders and admissible orders in the PF/IVPF contexts. This means that the proposed measures can overcome some drawbacks and ambiguities of the previous techniques through score and accuracy functions and can address incomparable PF/IVPF information more effectively. As an application of the risk attitude-based score functions, this paper develops novel risk attitudinal assignment models to establish a useful outranking approach for solving multiple criteria decision-making problems. Two algorithms for PF and IVPF settings are developed based on the concepts of a precedence frequency matrix and a precedence contribution matrix for conducting multiple criteria evaluation and the ranking of alternatives. A financing decision on aggressive/conservative policies of working capital management is presented to demonstrate the applicability of the proposed outranking approach in real situations. Moreover, a comparison to the technique of risk attitudinal ranking methods is investigated to validate the advantages of the proposed methodology. Furthermore, a comparative analysis with a newly developed outranking method, the IVPF-ELECTRE (for the elimination and choice translating reality), is conducted via an application to the investment problem of software development projects. Compared with these benchmark approaches, the proposed methods can produce a more reasonable and persuasive result for ranking the order relationships of alternatives in the highly uncertain context. The practical studies and comparative discussions demonstrate an excellent performance of the developed methodology that is effective and flexible enough to accommodate more-complex decision-making environments.

This paper presents an improved fingerprint recognition system for overlapping fingerprint of two fingers. The developed system can automatically and precisely separate an overlapping fingerprint into two areas, an overlapping area and a non-overlapping area, by analyzing their orientation fields and complicities. A method being able to remove a misjudging core of the fingerprint from the overlapping area is also proposed in this paper. Moreover, a recursive correction algorithm and a restrictions relaxation labeling algorithm are applied to separate two fingerprints from the determined overlapping fingerprint area, and the Gabor filter is applied to enhance figure quality of the two separated fingerprints. After that, the VeriFinger 6.2 SDK is used to identify the fingerprints. In the experiments, our system uses the fingerprints from the two public databases. Receiver Operating Characteristic (ROC) and Cumulative Match Characteristic (CMC), are used to examine the proposed fingerprint recognition system. The average recognition rate of our system is 85.7%. The experiments results show the effectiveness of the proposed system.

A parking lot (PL) dynamic resource allocation system for recharging electric vehicles (EVs) is introduced in this paper. For scheduling purposes, a day is divided into sequential timeslots. At the beginning of each timeslot, the dynamic system can determine an optimal charging schedule for that timeslot, as well as plan for subsequent timeslots. An EV may arrive at a PL with or without an appointment. Considering the variation in electricity prices during the day, the objective is to minimize the cost of electricity used to charge EVs by scheduling optimal electric quantities at the parking timeslots of each EV. The optimal solution satisfies the EV’s charging rate limit and the PL’s transformer limit. Based on particle swarm optimization (PSO), fuzzy systems and heuristics, this paper describes a heuristic fuzzy particle swarm optimization (PHFPSO) algorithm to solve the optimization problem. From the case studies, the results show the proposed dynamic resource allocation system has a significant improvement in satisfying charging requests and in reducing the electricity cost of the PL when compared with other scheduling mechanisms.

Nowadays, operation managers usually need efficient supply chain networks including important design factors such as economic and social considerations. The recent decade has seen a rapid development of controlling the uncertainty in supply chain configurations along with proposing novel solution approaches. By investigating the related studies, this paper shows that most of the current studies consider the economic aspects and just a few works present the two-stage stochastic programming as well as social considerations to design a closed-loop supply chain network. This motivated our attempts to consider economic and social aspects simultaneously by using the mentioned suppositions among the first studies. Another main contribution of this paper is the hybridization and tuning of a number of recent algorithms to address the problem. The results show that the proposed hybrid metaheuristic algorithms outperform the best existing techniques on the majority of case studies.

An unbounded external archive (UEA), which stores all nondominated solutions found during the search process, is frequently used to evaluate the performance of multi-objective evolutionary algorithms (MOEAs) in recent studies. A recent benchmarking study also shows that decomposition-based MOEA (MOEA/D) is competitive with state-of-the-art MOEAs when the UEA is incorporated into MOEA/D. However, a parameter study of MOEA/D using the UEA has not yet been performed. Thus, it is unclear how control parameter settings influence the performance of MOEA/D with the UEA. In this paper, we present an analysis of control parameters of MOEA/D under two performance evaluation scenarios. One is a final population scenario where the performance assessment of MOEAs is performed based on all nondominated solutions in the final population, and the other is a reduced UEA scenario where it is based on a pre-specified number of selected nondominated solutions from the UEA. Control parameters of MOEA/D investigated in this paper include the population size, scalarizing functions, and the penalty parameter of the penalty-based boundary intersection (PBI) function. Experimental results indicate that suitable settings of the three control parameters significantly depend on the choice of an optimization scenario. We also analyze the reason why the best parameter setting is totally different for each scenario., This is an accepted version of a paper published in Applied Soft Computing

Image semantic segmentation is more and more being of interest for computer vision and machine learning researchers. Many applications on the rise need accurate and efficient segmentation mechanisms: autonomous driving, indoor navigation, and even virtual or augmented reality systems to name a few. This demand coincides with the rise of deep learning approaches in almost every field or application target related to computer vision, including semantic segmentation or scene understanding. This paper provides a review on deep learning methods for semantic segmentation applied to various application areas. Firstly, we formulate the semantic segmentation problem and define the terminology of this field as well as interesting background concepts. Next, the main datasets and challenges are exposed to help researchers decide which are the ones that best suit their needs and goals. Then, existing methods are reviewed, highlighting their contributions and their significance in the field. We also devote a part of the paper to review common loss functions and error metrics for this problem. Finally, quantitative results are given for the described methods and the datasets in which they were evaluated, following up with a discussion of the results. At last, we point out a set of promising future works and draw our own conclusions about the state of the art of semantic segmentation using deep learning techniques. This work has been funded by the Spanish Government TIN2016-76515-R funding for the COMBAHO project, supported with Feder funds. It has also been supported by a Spanish national grant for PhD studies FPU15/04516 (Alberto Garcia-Garcia). In addition, it was also funded by the grant Ayudas para Estudios de Master e Iniciacion a la Investigacion from the University of Alicante.

In recent years, analysis and interpretation of video sequences to detect and track objects of interest had become an active research field in computer vision and image processing. Detection and tracking includes extraction of moving object from frames and continuous tracking it thereafter forming persistent object trajectories over time. There are some really smart techniques proposed by researchers for efficient and robust detection or tracking of objects in videos. A comprehensive coverage of such innovative techniques for which solutions have been motivated by theories of soft computing approaches is proposed. The main objective of this research investigation is to study and highlight efforts of researchers who had conducted some brilliant work on soft computing based detection and tracking approaches in video sequence. The study is novel as it traces rise of soft computing methods in field of object detection and tracking in videos which has been neglected over the years. The survey is compilation of studies on neural network, deep learning, fuzzy logic, evolutionary algorithms, hybrid and recent innovative approaches that have been applied to field of detection and tracking. The paper also highlights benchmark datasets available to researchers for experimentation and validation of their own algorithms. Major research challenges in the field of detection and tracking along with some recommendations are also provided. The paper provides number of analyses to guide future directions of research and advocates for more applications of soft computing approaches for object detection and tracking approaches in videos. The paper is targeted at young researchers who will like to see it as platform for introduction to a mature and relatively complex field. The study will be helpful in appropriate use of an existing method for systematically designing a new approach or improving performance of existing approaches.

In many engineering applications, experimental data is scant and tedious to generate. When performing a preliminary analysis of such data, the analyst is often challenged by the small size of the data set. An alternate, widely used approach to address this is bootstrapping, where synthetic samples are generated by randomly sampling the available dataset. In this paper, we suggest a way to augment small data sets in a manner that least distorts the original distributions. We do this by first projecting the dataset into the subspace spanned by the first k principal components, where k is appropriately chosen; next, additional samples are generated in the lower dimensional space. These additional samples are then projected back into the original space. The focus of this paper is on the inverse projection task, which arises in many other contexts. Well known solutions include using the Moore-Penrose inverse, which yields a minimum norm solution; an alternative is to solve a linear system of equations. However, these solutions do not typically preserve bounds on the input data. We propose a solution that allows a trade-off between minimizing the norm of the solution and the approximation error involved in the projection. We further develop a least-squares version of this approach in order to obtain a non-iterative procedure to obtain the inverse projection. We name our approach EigenSample, since additional samples are added in the eigenspace of the original dataset. The performance of our proposed approach is compared with three competing approaches for augmenting datasets and evaluated using multiple classification algorithms; and the results indicate the superior performance of our proposed method.

Collaborative filtering (CF) is widely applied in recommender systems to predict user preferences or interests according to a user’s historical information. Traditional CF methods mainly adopt batch-processing methods to train recommender models, which require prior preparation of all training data. However, user preferences or interests always change with time, and it is impossible to prepare all of the training data at once. In actuality, the data are obtained with time with a certain sequence. Therefore, to update the model, the trained model needs to be re-trained according to all of the historical datasets. As a result, the cost of re-training is very high, which slows down the recommender model, and makes new updates of the user data difficult to capture. To solve these problems, the online recommender system emerged. In this paper, we propose a confidence-weighted bias model (CWBM) for online collaborative filtering (OCF). This model adds bias into CF and further introduces confidence weights; thus, it can improve the stability and accuracy of OCF. A comparative experiment on two real datasets, Movielens100K and Mvielens1M, show that the method proposed in this paper is superior to other baseline methods.

This paper considers two-agent scheduling problem with a single machine which is responsible for processing jobs from two agents. The objective is to minimize the objective function of one agent, subject to an upper bound on the objective function of the other agent. The objectives considered in this paper are, (1) the minimization of total completion time and (2) the minimization of total weighted completion time. To solve these problems, one heuristic and an Ant Colony Optimization algorithm are proposed. The heuristic suggested in the paper are motivated by the Weighted Shortest Processing Time first (WSPT) rule. A numerical experiment is performed on randomly generated problem instances. The performance of the algorithm is evaluated by comparing it with the lower bound value of all three problems considered in the present paper.

This paper presents an efficient non-iterative hybrid empirical mode decomposition (EMD) and kernel ridge regression (KRR) for significantly accurate short-term wind speed and wind power prediction. The original non-linear and non-stationary wind speed and wind power time series data are decomposed using EMD to isolate the mutual effects between different components. The proposed EMD-KRR model is tested for predicting wind speed and wind power time series data over a time horizon spanning intervals of 10 min, 30 min, 1 h and 3 h ahead, respectively. Further the performance of EMD-KRR prediction model is compared with two other widely used non-iterative prediction models like EMD based Random vector functional link network (RVFL), and EMD based Extreme Learning Machine (ELM). Also an iterative Mutated Firefly Algorithm with Global Optima concept (MFAGO) optimized RVFL network is used for comparison to prove the advantages of non-iterative models over the iterative ones. The performance metrics of the proposed EMD-KRR and EMD-RVFL confirm the effectiveness and precision in producing an accurate forecast of both wind speed and wind power in comparison to all other prediction models using the wind power data from three real world wind farms. Further a fast reduced version of the EMD-KRR is presented in the paper to reduce the computational overhead substantially using randomly selected support vectors from the data set while resulting in a reasonably accurate forecast.

A novel method is proposed in this paper to extend the traditional artificial bee colony algorithm to competently optimize noisy objective function. The first strategy is concerned with adaptive selection of sample size for periodic evaluation of noisy objective function of a trial solution to reduce the possible risk of promoting noisy solutions over generations. The sample size of a given solution here is adapted based on the level of noise contamination in its local neighborhood. In this paper, the noise distribution in the locality of a candidate solution is captured by its local neighborhood fitness variance. The paper also recommends a new stratagem to identify local neighborhoods of trial solutions of a noisy optimization problem. The policy to adapt sample size here is realized using stochastic learning automata. The second strategy deals with evaluating the effective fitness of a trial solution based on a measure of relative distance between its noisy fitness samples. The adopted policy helps to capture the uncertainty in measure of fitness samples of a given solution due to noise contamination. Unlike conventional averaging, the proposed strategy aims at giving primacy to the true samples than the noisy samples of a trial solution during estimating its effective fitness. Last, a probabilistic selection strategy is introduced to diminish the jeopardizing effect of noise during deterministic selection of trial solutions to form next generation population. Computer simulations undertaken on the noisy versions of 38 benchmark functions reveal that the proposed algorithm surpasses its competitors with respect to function error value.

This paper aims to provide a novel metaheuristic algorithm motivated from quantum motion entitled Multi-scale Quantum Harmonic Oscillator Algorithm (MQHOA). The calculation accuracy of MQHOA is adjustable. The physical model and mathematical analysis of the proposed algorithm are detailed and well interpreted in this paper. The structure of MQHOA is very simple, including merely two phases, the quantum harmonic oscillator process (QHO process) and multi-scale process (M process). Experiments are carried out to validate the effectiveness and efficiency of MQHOA by applying it to 30 well defined benchmark functions. We also compare MQHOA with several well-known metaheuristic algorithms, such as genetic algorithm (GA), simulated annealing (SA), particle swarm optimization (PSO) and quantum particle swarm optimization (QPSO). The comparative results indicate the competitive and superior performance of the proposed algorithm in both convergence speed and optimal solution accuracy.

Multisource assessment (MSA) is a common employee performance evaluation model in which subjective views of different groups of assessors about individuals are collected and aggregated. Since in the MSA model, uncertain and subjective assessments are usually aggregated using the Averaging method, the quality of its outputs has been questioned in many studies. Regarding the importance of the data aggregation problem and lack of a proper solution in this context, in this paper, a new model based on the Evidence Theory has been proposed to improve data aggregation process in MSA model. To determine the best implementation mode, the performance of the proposed model has been investigated in various defined conditions through a simulation study. The findings of the paper reveal that, in comparison with the traditional aggregation method, the proposed model significantly increases the accuracy of the outputs in the MSA model.

Sparse coding is based on the concept of having a large dictionary of candidate basis vectors. Any given vector is expressed as a sparse linear combination of the dictionary vectors. It has been developed in the signal processing field, and has many applications in data compression and image processing. In this paper we propose applying sparse coding to the time series forecasting field. Specifically, the paper investigates different dictionary based local learning techniques for building predictive models for the time series forecasting problem. The proposed methodology is based on a local learning framework whereby the query point is embedded and coded in terms of a sparse combination of the training dictionary atoms (vectors). Then this embedding is used for estimating the target value of the query point, by applying the same embedding to the target vectors of the dictionary training atoms. We present an experimental study of several sparse coding algorithms. Experiments are performed on the large monthly time series benchmark from the M3 competition, and these experiments showed that the sparse methods Lasso and Elastic-Net presented the best results among the sparse coding algorithms. Moreover, they outperformed the K-nearest neighbor (KNN) regression and most of the compared machine learning and statistical forecasting techniques, especially for higher horizons.

In this paper, a new hierarchical method of contour analysis is proposed. The first stage of the analysis is based on the syntactic approach in which elementary segments of the contour are extracted by using geometric features. As a result, each extracted segment belongs to one of the equivalence classes which make the set of primitives. The contour is described as a string of primitives. The applied syntactic approach is a generalization of a classical shape language. Such description enables to classify some cases, but not all. Therefore, the second stage of the classification is applied. It consists in analysing other local geometric features of the contour in the context of the syntactic description obtained at the first stage of the analysis. The values of these features are used to define the arguments of the membership functions of the family of fuzzy sets which is the basis for the contour classification at the second stage of the analysis. The proposed formalism has been applied effectively to the recognition of bone contours in X-ray images. Healthy bones, the bones with erosions and the bones with osteophytes have been classified correctly. The proposal described in this paper turned out to be effective in the classification of contours of bones.

In this paper, an iterated local search algorithm is proposed to find the feasible solution for the University Course Timetabling Problem. Three key phases are involved in the proposed algorithm framework: initialization, intensification and diversification. Once a partial-feasible initial timetable is constructed, a simulated annealing based local search and a diversification procedure that brings moderate perturbation or even improvement to the current solution are performed in an iterative manner until a stop condition is met. The proposed algorithm is evaluated on a widely used dataset containing 60 problem instances. The computational results show the iterated local search algorithm achieves highly competitive results compared with the existing algorithms. It is noteworthy that this algorithm can find feasible solutions for 58 instances in reasonable time, including three large instances whose feasible solutions are missed in previous papers. Furthermore, some key elements and properties of the algorithm are also analyzed.

The extensive use of hazardous products has resulted in quickly increasing on hazardous wastes. Due to the rising environmental pressure and economic benefit, the reverse supply chain design for hazardous products is becoming increasingly important and urgent. In this paper, we consider the closed-loop supply chain network design for hazardous products (HP-CLSCND), including both forward supply chain and reverse supply chain. The uncertainty inherent in closed-loop supply chain network will significantly influence the overall performance of the closed-loop supply chain network design. This paper focuses on the HP-CLSCND problem with uncertain demands and returns, and a two-stage stochastic programming model (scenario-based) is proposed, in which a risk restriction constraint and reward-penalty mechanism are simultaneously considered. Two solution approaches, parallel enumeration method (PEM) and genetic algorithm (GA) are designed to solve the proposed model. The PEM is an exact solution approach and can rapidly obtain the global optimal solution of the proposed model by utilizing multiply processors. Finally, an application example is provided to demonstrate the applicability of the proposed model and two solution approaches. The performance of PEM is evaluated by speedup radio. In addition, the sensitivity analyses about maximum acceptable risk and reward-penalty intensity are conducted, and some management insights for the government are obtained.

In this paper we present a novel approach for solving the consequent part of neuro-fuzzy modeling with an emphasis on the forgetting factor in the multi-class learning problem. Our solution is based on recursive least squares (RLS) for online and incremental learning applications, where the data stream is not necessarily uniformly distributed over time. Such a setup can lead to forgetting of specific classes that have not been used for a period of time. In this work we present a reasoned and detailed description of elastic memory learning (EML) and EML with the use of confidence interval (EML+) to avoid unnecessary treatment of the forgetting factor. We present the experimental results and evaluation of our methods in order to show their usefulness not only against forgetting of unused classes, but also for dealing with the lowered recognition rate after all classes have been learned. We note that by using EML, forgetting is significantly eliminated and the recognition rate is slightly affected as well, while EML+ puts more emphasis on keeping the recognition rate higher than the forgetting. Thus, this paper presents two methods that significantly eliminate the forgetting factor for incremental learning with a different focus on its importance, i.e. high recognition rate vs high immunity to forgetting; both of these methods perform significantly better than RLS for these aspects.

Venture capitalists are confronted with decision-making problems under an uncertain investment circumstance, which lead them to rely heavily on intuition and psychological state to make their decisions in most cases. It has attracted more and more researches to focus on such behaviors of venture capitalists in recent years and along with the development of behavioral finance in venture capital field. Through the review of framework of venture capitalists’ decision-making and the investigation of background of venture capital market in China, we find that the decision-making framework, which is based on the traditional expected utility theory with hypothesis that the investor is completely rational, is inconsistent with reality. Moreover, the decision-making information is often depicted as crisp number with probability that couldn’t express the uncertainty which is the fundamental nature of venture capital circumstance. Therefore, a novel approach that considers both the bounded rationality of venture capitalists and the uncertain circumstance of venture capital in decision-making is urgent to select the promising enterprise in China. Fortunately, prospect theory, as one of the greatest achievement in behavioral finance, thinks that venture capitalists focus on subjective probabilities instead of objective probabilities and show the behavior characterized by risk aversion over gains and risk seeking over losses, which can depict the psychological state of venture capitalists properly. Furthermore, the intuitionistic fuzzy information can express essential uncertainty in venture capital circumstance accurately. Therefore, we adopt Intuitionistic Fuzzy Prospect Theory that integrates both the merits of prospect theory and intuitionistic fuzzy information to aid venture capitalists to make a better decision in the real world. In this paper, the detailed steps of how to select a promising enterprise has been illustrated. Then, a practical example of selecting the promising enterprise for venture capitalists in Ali Capital is provided to exhibit the process of application and exemplify the feasibility of this selecting approach. Meanwhile, from the comparison of ranking results derived by TOPSIS and this approach, we find that the approach proposed in this paper is superior to the TOPSIS in selecting the promising enterprise according to the subsequent funded results of enterprises.

Multiple Criteria Decision-Making (MCDM) methods are widely used in research and industrial applications. These methods rely heavily on expert perceptions and are often sensitive to the assumptions made. The reliability and robustness of MCDM analysis can be further tested and verified by a computer simulation and sensitivity analysis. In order to address this, five different MCDM approaches, including Weighted Sum Model (WSM), Weighted Product Model (WPM), revised Analytic Hierarchy Process (rAHP), Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) and COmplex PRoportional ASsessment (COPRAS) are explored in the paper. Real data of the case study for assessing housing affordability are used for testing the robustness of alternative ranking and finding the most sensitive criteria to the change of criterion weight. We identify the most critical criteria for any and best ranking alternatives. The paper highlights the significance of sensitivity analysis in assessing the robustness and reliability of MCDM outcomes. Furthermore, randomly generated and model-based data sets are used to establish relationship between the dispersion of relative importance values of alternatives and ranking uncertainty. Our findings demonstrate that the dispersion of relative importance values of alternatives correlate with the Euclidian distances of aggregated values. We conclude that the dispersion of relative importance values contributes directly to the ranking uncertainty and can be used as a measure for identifying critical criteria.