Search

With the advancement of machine learning and radar technology, machine learning is becoming more and more widely used in the field of radar. Radar scanning, signal acquisition and processing, one-dimensional range image, radar SAR, ISAR image recognition, radar tracking and guidance are all integrated into machine learning technology, but machine learning technology relies heavily on human machine learning experts for radar signal recognition. In order to realize the automation of radar signal recognition by machine learning, this paper proposes an automatic machine learning AUTO-SKLEARN system and applies it to radar radiation source signals. Identification: Firstly, this paper briefly introduces the classification of traditional machine learning algorithms and the types of algorithms specifically included in each type of algorithm. On this basis, the machine learning Bayesian algorithm is introduced. Secondly, the automatic machine learning AUTO based on Bayesian algorithm is proposed. -SKLEARN system, elaborates the process of AUTO-SKLEARN system in solving automatic selection algorithm and hyperparameter optimization, including meta-learning and its program implementation and automatic model integration construction. Finally, this paper introduces the process of automatic machine learning applied to radar emitter signal recognition. Through data simulation and experiment, the effect of traditional machine learning k-means algorithm and automatic machine learning AUTO-SKLEARN system in radar signal recognition is compared, which shows that automatic machine learning is feasible for radar signal recognition. The automatic machine learning AUTO-SKLEARN system can significantly improve the accuracy of the radar emitter signal recognition process, and the scheme is more reliable in signal recognition stability.

Communities in social networks are the essential feature which may be considered as a potential parameter in modeling the behavior of the social entities. Detection of communities has attracted a lot of attention in research in social network analysis. It is one of the major challenging problems as it involves high complexity in processing complex web structure. In fact, this problem can be considered as a NP-complete problem in large-scale networks, as this problem is somewhat reducible to the clique problem in graph theory. A number of meta-heuristic algorithms have been proposed to explore the hidden communities. Most of these algorithms have considered the modularity of the network as their objective function. But, the aspect of optimizing the value of modularity is associated with a problem known as resolution limit, where the size of the detected communities depends on the number of edges existing in the network. In this paper, a genetic algorithm-based community detection has been proposed where an efficient single objective function based on similarity matrix has been devised. The similarity index between each pair of nodes has been calculated in a distributed manner over multiple computing nodes. Similarity index proposed in this paper is based on the topological structure of the network. The effectiveness of the proposed approach is examined by comparing the performance with other state-of-the-art community detection algorithms applied over some real-world network datasets.

A novel extreme-learning-machine-based robust control scheme for automotive electronic throttle systems with uncertain dynamics is presented in this paper. It is shown that the well-known extreme learning machine (ELM) is used to estimate the upper bound of the lumped uncertainty while a fast nonsingular terminal sliding mode feedback controller is designed to achieve global stability and finite-time convergence for the closed-loop system. Although the ELM used in this paper has the same structure as the one in the conventional least-square-based ELM used for pattern classifications, i.e., the input weights are randomly chosen, the ELM adopted in the closed-loop control system is designed to achieve global control purpose. The output weights of the ELM will be adaptively adjusted in Lyapunov sense from the perspective of global stability of the closed-loop system, rather than local optimization in conventional ELM. The proposed control can thus not only realize the finite-time error convergence but also needs no prior knowledge of lumped uncertainty. Simulation results are demonstrated to verify the excellent tracking performance of the proposed control in comparison with other existing control methods.

In this paper, we investigate a multi-objective multi-item fixed-charge solid transportation problem (MOMIFCSTP) with fuzzy-rough variables as coefficients of the objective functions and of the constraints. The main focus of the paper is to analyze MOMIFCSTP under a fuzzy-rough environment for a transporting system. In practical situations, the parameters of a MOMIFCSTP are imprecise in nature, due to several uncontrollable factors. For these reasons, we introduce the fuzzy-rough variables in MOMIFCSTP to tackle vague data which are different from fuzziness and roughness. Fuzzy-rough expected-value operator is employed to convert fuzzy-rough MOMIFCSTP into deterministic MOMIFCSTP. Thereafter, we develop a methodology to solve the deterministic MOMIFCSTP by technique for order preference by similarity to ideal solution (TOPSIS). Three distinct approaches, namely extended TOPSIS, weighted goal programming (WGP) and fuzzy programming, are used to derive Pareto-optimal solution from the suggested model. A comparison is drawn among the optimal solutions which are derived from different approaches. It is observed from the extracted results that TOPSIS provides a better optimal solution than WGP and fuzzy programming. TOPSIS also overcomes some difficulties which arise in WGP. Finally, a real-world (industrial) problem is incorporated to show the applicability and feasibility of the proposed problem.

Today’s industry demands precise functioning and zero failure of rotating machinery (RM) to avoid disastrous accidents as well as financial losses. Rolling element bearings (REBs) are the heart of RM. Therefore, as early as possible to provide the significant time for maintenance planning, an intelligent diagnosis of REB fault is a critical and challenging task. Thus, this paper presents an efficient method for fault diagnosis. The proposed method mainly consists of two consecutive units: (1) generation of kurtogram of raw vibration signal and (2) training of extreme learning machine (ELM) classifier using kurtogram. Kurtogram has a distinct capability to represent the hidden non-stationary components of a raw signal. Therefore, it is considered as a unique feature vector for fault classification. ELM is a well-organized fast learning method proposed by Huang et al. and showed that it is better than traditional learning algorithms. However, one of the open issues of ELM is to design compact-size ELM architecture by preserving the accuracy of the solution. Thus, improved random increment ELM is proposed in this paper. Initially, it randomly adds the nodes to network architecture to rapidly reduce the residual error up to the predefined threshold and then sequentially adds the nodes to the network architecture for further reducing the residual error. Performance of the proposed routine is evaluated by REB vibration data: artificially generated vibration data and Case Western Reserve University bearing data. The experimental study reveals the classification accuracy of the proposed approach with both the datasets for various faults and also compared with existing methods.

In the big data environment, hospital medical data are also becoming more complex and diversified. The traditional method of manually processing data has not been able to meet the management needs of massive medical data. With the further development of big data technology and machine learning, the smart medical aided diagnosis model came into being. However, there is almost no auxiliary diagnosis mode for orthopedic diseases. In order to make up for the gap in the auxiliary diagnosis of orthopedics and promote the wisdom process of orthopedic disease diagnosis, this paper proposes an orthopedic auxiliary classification prediction model based on XGBoost algorithm. The experimental data were obtained from the clinical case information of femoral neck patients from April 2016 to October 2018, Department of Bone and Soft Tissue Tumor Surgery, Cancer Hospital of China Medical University, Liaoning Cancer Hospital & Institute. In order to make the experimental results more convincing, while constructing the XGBoost model, the orthopedic auxiliary classification prediction model is constructed based on the random forest algorithm and the associated classification algorithm, respectively, and the three models are compared and analyzed. The results show that compared with the random forest model and the associated classification model, the XGBoost algorithm classification prediction model has higher accuracy, faster calculation speed, and more applicability in orthopedic clinical data. The XGBoost algorithm can cope with complex and diverse medical data, and can better meet the requirements of timeliness and accuracy of auxiliary diagnosis. The classification and prediction model of orthopedic auxiliary diagnosis proposed in this paper helps to reduce the workload of medical workers, help patients prevent and recover early, and realize real auxiliary medical services.

This paper highlights the ability of convolutional neural networks (CNNs) at classifying EEG data listening to different kinds of music without the requirement for handcrafted features. Deep learning architectures presented in this paper include CNN of different depths and different convolutional kernels. Support vector machine (SVM) taking in EEG features describing the frequency spectrum, signal regularity, and cross-channel correlation has been applied for performance comparison with CNN. The best performing CNN model presented in this paper achieves the tenfold cross-validation (CV) binary classification average accuracy of 98.94% (validation) and 97.46% (test), and the tenfold CV three-class classification accuracy of 97.68% (validation) and 95.71% (test). In comparison, the SVM classifier achieves tenfold CV binary classification accuracy of 80.23% (validation). The CNN model presented is able to not only differentiate EEG of subjects listening to music from that of subjects without auditory input, but it is also capable of accurately differentiating the EEG of subjects listening to different music. In the context of designing neural computing models for EEG analysis, this paper shows that decomposing two-dimensional spatiotemporal convolutional kernels into separate one-dimensional spatial and one-dimensional temporal kernels significantly reduces the number of trainable parameters (size) of the model while retaining the classification performance. This finding is useful, especially in designing CNN for memory-critical embedded systems for EEG processing. In neurological aspect, auditory stimulus is found to have altered the EEG pattern of the frontal lobe and the left cerebral hemisphere more than the other brain regions.

Computer vision technology is used to analyze X-ray images and detect dangerous goods in the process of logistics and express delivery. It is a security technology which can reduce labor strength and improve working efficiency. At present, there are many excellent detection models and methods in the field of object detection for visible light images, such as R-CNN, Fast R-CNN, Faster R-CNN, YOLO, SSD. These deep neural network-based detection methods achieved excellent performance on ImageNet. The training of object detection models on X-ray image datasets for dangerous goods detection is the focus of research in the field. Due to practical reasons, it is difficult to collect a comprehensive image dataset of dangerous goods (positive samples). In order to overcome this problem, this paper uses a multi-task transfer learning method on the basis of classification task and location search task on SSD network. The research in this paper focuses on adding additional convolutional layers in the SSD network to re-learn the knowledge of the model learned from the source domain. Experiments show that compared with the traditional method of fine-tuning, this method has better transfer learning ability on SSD network. This method was used to perform experiments in SSD300 on the image datasets screened from GDXray and achieved a mean average precision (mAP) of 0.915.

At present, there are problems such as low level of drug inventory management in major hospitals, which directly lead to the stagnation of the drug supply chain and the high cost of hospital management. In order to improve the efficiency of hospital drug inventory management, based on genetic algorithm and BP neural network, this study combined the actual situation of hospital drug inventory forecast to build a system model based on hospital drug management mode. At the same time, in order to verify the validity of the model, this paper collected experimental data through data collection, inputted data for model simulation analysis, and established a three-layer network simulation model. In addition, this paper analyzed the accuracy and sensitivity of model prediction based on actual conditions and compared the research results with actual data. Studies have shown that the model proposed in this study has certain effects and can be applied to hospital drug inventory management practices.

Considering the influence of carbon emissions trading, the fuzzy stochastic programming model was established to cut back the total cost of carbon trading balance. Modeling this chain is carried out by accounting for carbon cap-and-trade considerations and total cost optimization. In this paper, we analyze the low-carbon integrated forward/reverse logistics network and made relevant simulation tests. The results show that the changes of the confidence level and carbon emission limits have obvious influences on logistics costs. If the emission limit is large, carbon trading mechanism has little effect on the total logistics cost in the same scenario. Therefore, the government needs to use the appropriate emission limits to guide enterprises to reduce carbon emissions, and enterprises can make coping strategies according to the different limit at the same time. Therefore, the fuzzy random programming model proposed in this paper is practical. Its decision making applying the proposed algorithm is reasonable and applicable and could provide decision basis for enterprise managers.

Instability is a common problem in deep belief network–back propagation forecasting model, and the trend of traffic data will affect the forecasting results of the model. Therefore, this paper proposes a short-term traffic flow forecasting method based on deep belief network–support vector regression. Support vector regression classifier SVR is used at the top of the model. Data processing is from bottom to top. Firstly, at the bottom of the model, the input traffic flow data are processed differently; then, the DBN model is used to learn the traffic flow characteristics. Finally, SVR is used to predict the traffic flow at the top of the model. The average absolute error of the prediction is 9.57%, and the average relative error is 5.91%. The relationship between the predicted value and the actual traffic flow data is found through simulation experiments. The predicted value of the model proposed in this paper is in good agreement with the measured value, and the prediction accuracy is high. The model can effectively predict short-term traffic flow. Finally, compared with the traditional DBN prediction model and other common prediction models, the proposed prediction model has higher prediction accuracy.

This paper extends one of the most extensively used multiple criteria decision analysis (MCDA) methods, the technique for order preference by similarity to ideal solutions (TOPSIS), to adapt to highly complicated uncertain environments based on interval-valued Pythagorean fuzzy (IVPF) sets. In contrast to classical TOPSIS methods, this paper develops a novel concept of Pearson-like IVPF correlation coefficients instead of distance measures to not only construct a useful and effective association measure between two IVPF characteristics but also depict the outranking relationship of IVPF information. Moreover, this paper proposes the (weighted) IVPF correlation-based closeness coefficients to establish a Pearson-like correlation-based TOPSIS model to manage MCDA problems within the IVPF environment. In particular, there is a definite improvement in determining the closeness coefficient required in the TOPSIS procedure. This paper considers anchored judgments with respect to the positive- and negative-ideal IVPF solutions and provides new approach- and avoidance-oriented definitions for the IVPF correlation-based closeness coefficient, which is entirely different from the traditional definition of relative closeness in TOPSIS. Furthermore, this paper proposes a comprehensive IVPF correlation-based closeness index to balance the consequences between ultra-approach orientation and ultra-avoidance orientation and acquire the ultimate compromise solution for decision support and aid. The feasibility and practicability of the developed methodology are illustrated by a practical MCDA problem of rehabilitation treatment for hospitalized patients with acute stroke. The application results, along with experimentations and comparative analyses, demonstrate that the developed methods are rational and effective.

The aim of this paper is to give a multi-attribute decision-making (MADM) method for the interval-valued intuitionistic fuzzy (IVIF) set using the set pair analysis (SPA) theory. IVIF set can express the uncertain information in a more valuable manner, while the connection number (CN) based on the “identity,” “discrepancy” and “contrary” degrees of the SPA theory handles the uncertainties and certainties systems. Based on these features, in this paper, we develop a new possibility degree measure based on the CNs to rank the different IVIF numbers. The properties and advantages of the measure are described in details. Later on, we present a novel MADM method and illustrate with several examples to validate it. A comparative as well as superiority analysis is performed to show the feasibility and validity of the approach.

A new hybrid model combining the kernel functions along with the random vector functional link neural network (RVFLN) is proposed in this paper for an effective solar power prediction. The conventional RVFLN is known for its fast learning speed, simple architecture and good generalization capabilities and allows direct connection between input and output nodes along with nonlinear enhancement nodes with random weights. However, the bottleneck of selecting the number of hidden enhancement nodes and mapping functions is still a challenging problem. To overcome these deficiencies of the conventional RVFLN, kernel functions are used for both the direct links and the hidden nodes to provide better stability, generalization and regression accuracy. Instead of using a single kernel for the enhancement nodes, this paper proposes an optimal kernel function that comprises a linear combination of weighted local kernel and a global kernel to improve the prediction accuracy of the solar power generation. This optimal kernel will be known as multi-kernel RVFLN (MK-RVFLN), and its parameters are optimized using an efficient metaheuristic evaporation-based water cycle (EVWCA-MKRVFLN) to provide accurate prediction of solar power. To validate its superior prediction accuracy, two solar power plants of 25 and 100 MW capacity in the states of New York and California are considered for 5-min- and 60-min-ahead prediction in the months of January, April, July and October. The result analysis shows that the MK-RVFLN algorithm attains better performance than many other techniques.

To describe the system fault evolution process, the space fault network theory is proposed. Space fault network theory is the third stage of space fault tree theory. The paper introduces the basic research results of space fault network. The basic ideas, basic definitions and corresponding physical meanings of space fault network are discussed. The properties, structure and space fault tree transformation method of space fault network are further studied. The characteristics of the system fault evolution process and its four elements are discussed. The representation methods of space fault network for system fault evolution process are presented. The transformation methods of the general space fault network and multidirectional ring space fault network into space fault tree are given and studied. In particular, the classification and characteristics of unidirectional ring space fault network are studied in detail. Based on the system fault evolution process of the example, we conducted a qualitative analysis, and a space fault network is established and transformed. At the same time, a quantitative analysis is carried out according to the derivation process of the transformation of multidirectional ring space fault network. The results show that the space fault network can effectively describe and analyze the system fault evolution process. This paper has solved some basic problems of system fault evolution process, but more complicated situations need further research.

Nowadays, digitizer pens have become front end of many digital devices. The increasing use of this technology has necessitated the need for producing pen-based virtual keyboard systems. Despite attempts to create such systems in English, their absence for Persian/Arabic languages is an obvious defect. The goal of this paper is presenting an online continuous Persian/Arabic character recognition method. A character in Persian/Arabic language is made of two types of signs or strokes: main body and delayed strokes (which may be zero or more sign). In this paper, a set of novel and discriminative spatial features are defined for these strokes. These features then are used in a novel algorithm to create a genetic programming-based decision tree called GPDT. The GPDT and spatio-temporal features are utilized by non-deterministic finite automata (NDFA) to recognize group-related strokes and related characters. The reason for using spatio-temporal features is the sameness of the main body of some Persian/Arabic letters (e.g., “ح، خ، ج، چ”). There are also two other issues related to recognizing Persian/Arabic letters: unknown number of delayed stroke segments and the sameness of delayed strokes placement, which are removed by using an NDFA. In fact, after identifying group of main body with the help of GPDT, each recognized stroke makes a move in NDFA to stop in a character state (final state on the end of a path in NDFA). The proposed algorithm recognizes continuous Persian/Arabic letters and digits with a 92.43% accuracy and isolated letters and digits with 97.52% accuracy.

Parameter settings for nature-inspired optimization algorithms are essential for their effective performance. Evolutionary algorithms and swarm intelligence algorithms are prominent types of nature-inspired optimization. There are comprehensive reviews of parameter setting techniques for evolutionary algorithms. Counterparts providing an overview of parameter setting techniques for swarm intelligence algorithms are needed also. Therefore, in this paper, we provide a critical and comprehensive review, focusing in particular on dynamic parameter setting techniques. The paper describes a variety of swarm intelligence algorithms and parameter setting approaches that have been applied to them. This review simplifies the selection of parameter setting techniques for each algorithm by collecting them in a single document and classifying them under a taxonomy. Recommendations for parameter setting approach selection are provided in this review. We explore the open problems related to dynamic parameter setting techniques for swarm intelligence optimization and discuss the trade-off between run-time computation and flexibility of these algorithms.

Evolutionary algorithms (EAs) have been successfully applied to solve numerous optimization problems. Neighborhood field optimization algorithm (NFO) has been proposed to integrate the neighborhood field in EAs, which utilizes local cooperation behaviors to explore new solutions. In this paper, certain new NFO variants are proposed based on the cooperation of descendent neighbors. The competitive and cooperative behaviors of NFO variants provide a remarkable ability to accelerate information exchanges and achieve global search. Experimental results show that NFO variants perform better than basic and other state-of-the-art EAs under different benchmark functions. For NFO and other EAs, it is difficult to quantify benefits of local cooperation in the optimization process. For this purpose, the cooperation behaviors are analyzed in a new network approach in this paper. In the proposed NFO variants, population graph shows a scale-free network with power-law distribution. Network characteristics, i.e., degree distribution, cluster coefficient and average degree, are used to quantify the cooperation behaviors. Experimental results show that network characteristics can effectively indicate the optimization performance of NFO variants in terms of convergence rate and population diversity. NFO variants with large cluster coefficients and significant heterogeneous characteristics can achieve a significant performance improvement on numerous problems.

With the rapid rise of the Internet, China’s e-commerce has also flourished. The development of e-commerce has led to an increase in the volume of logistics and distribution. The further development of e-commerce has also placed higher demands on the timeliness of logistics and distribution. The competition of e-commerce companies has shifted from the competition between business models to the competition between logistics services. The scientific and rational distribution site selection planning is the prerequisite and guarantee for the efficient operation of logistics distribution network. To balance the contradiction between logistics distribution speed and distribution cost has become the key to competition among e-commerce companies. This paper analyzes the current network structure and distribution mode of e-commerce logistics city distribution, and analyzes and discusses the problems existing in current e-commerce logistics city distribution. Furthermore, the bi-level programming is studied. According to the characteristics of the bi-level programming problem, the genetic algorithm flow suitable for bi-level programming is proposed. The bi-level programming model of urban distribution service network site selection with limited lines is proposed. Through the verification of the genetic algorithm in this paper, the proposed method can plan a reasonable service site location layout and distribution models and path selection. The results show that the average daily fuel cost can be reduced by 37.6%, and the transportation distance and fuel cost can be optimized best.

Gait has been well known as an unobtrusive promising biometric to identify a person from a distance. However, the effectiveness of silhouette-based approaches in gait recognition is diluted due to variations of view angles. In this paper, we put forward a novel and effective method of gait recognition: cross-view gait recognition based on ensemble learning. The proposed method greatly enhances the effectiveness and reduces the sensitivity of gait recognition under various view angles conditions. Furthermore, in this paper we will introduce a novel algorithm based on ensemble learning for combining several gait learners together, which utilizes a well-designed gait feature based on area average distance. Through experimental evaluations on the well-known CASIA gait database and OU-ISIR gait database, our paper demonstrates the advantages of the proposed method in comparison with others. The contribution of this research work is to resolve the multiview angles problem of gait recognition through assembling several gait learners.

The file attached to this record is the author's final peer reviewed version. The Publisher's final version can be found by following the DOI link. In conventional data envelopment analysis (DEA), a production system has been seen as a black box for measuring the efficiency without any attention to what is happening inside the system. However, in practice, performance improvement often requires observing the internal structure of the producing system in order to find the sources of inefficiencies. In addition, weight flexibility as a key property of the multiplier DEA models allows a system to totally disregard an assessment factor, either input or output, from the evaluation process by assigning a value of zero or epsilon to its weight. This paper contributes to the existing literature by proposing a common-weights DEA model when the production system includes a number of interrelated processes. To this end, we propose an aggregate DEA model to calculate the most favourable common weights for determining the efficiency of all production systems and their processes at the same time. Our proposed aggregate model not only is linear for equitably evaluating the producing units on the same scale, but also enables us to deal with the mixed network structures. Furthermore, the network system is decomposed into a series system to build a relational network DEA model that emphasises separate relatedness. This greatly reduces the computational complexities for enormous volumes of data in many real applications and treat difficulties in network DEA models including the zero value and fluctuating weights and multiple solutions. Managerially speaking, this paper aims to provide the top management team of a production system with an integrated framework to shape a better strategic decision process about firm performance, which is to treat the sources of inefficiencies and ultimately take corrective actions over the long run. Put differently, the proposed framework helps top managers make proper decisions in complex situations with a view of improving a firm’s efficiency in all production divisions, which can be identified as a core competency leading to competitive advantages of the organisation. In the context of performance management, our study is equipped with a simple numerical example and a case study of the non-life insurance companies to demonstrate the applicability of the proposed common-weights network model.

The bottom hole pressure (BHP) of managed pressure drilling (MPD) is a typically unstable object with hysteresis that is difficult to be directly controlled. However, at the present stage, BHP control still focuses on conventional PID control and simple intelligent control, requiring repeated data alignment. There are some related problems, such as lack of control over BHP, longer working hours and high cost of drilling. In order to increase economic effects of MPD, this paper analyzes the MPD system and utilizes wellhead back pressure as the controlled variable. According to throttle valve features, basic parameters and boundary conditions of MPD, a mathematical model of throttle valve is also calculated. Besides, this paper focuses on studying the control model and proposes an improved particle swarm algorithm to optimize PID neural network (IPSOPIDNN) model. This model is improved based on inertia weight and fitness function of conventional particle swarm algorithm. Moreover, the particle swarm algorithm is used to optimize the initial weight value of PID neural network, shorten the search time for optimal value of particle swarm, and reduce the chance of local minimum. The real-time control results of IPSO-PIDNN are compared with results of traditional particle swarm optimization PID neural network (PSO-PIDNN) and particle swarm optimization PID neural network (PSO-PIDNN). IPSO-PIDNN control system has some advantages, including favorable self-learning, optimization quality, high levels of control precision, no overshoot, rapid response and short setting time. In this way, advanced automation control of BHP is conducted during managed pressure drilling process, thus providing technical support for the well control safety of managed pressure drilling.

Recent growth of metaheuristic search strategies has brought a huge progress in the domain of computational optimization. The breakthrough started since the well-known Particle Swarm Optimization algorithm had been introduced and examined. Optimization technique presented in this contribution mimics the process of flower pollination. It is build on the foundation of the first technique of this kind—known as Flower Pollination Algorithm (FPA). In this paper, its simplified and improved version, obtained after extensive performance testing, is presented. It is based on only one natural phenomena—called flower constancy—the natural mechanism allowing pollen carrying insects to remember the positions of the best pollen sources. Modified FPA, named as Biotic Flower Pollination Algorithm (BFPA) and relying solely on biotic pollinators, outperforms original FPA, which itself proved to be very effective approach. The paper first presents a short description of original FPA and the changes leading to Biotic Flower Pollination Algorithm. It also discusses performance of the modified algorithm on a full set of CEC17 benchmark functions. Furthermore, in that aspect, the comparison between BFPA and other optimization algorithms is also given. Finally, brief exemplary application of modified algorithm in the field of probabilistic modeling, related to physics and engineering, is also presented.

The recognition of staining patterns present in human epithelial type 2 (HEp-2) cells helps to diagnose connective tissue disease. In this context, the paper introduces a robust method, termed as CanSuR, for automatic recognition of antinuclear autoantibodies by HEp-2 cell indirect immunofluorescence (IIF) image analysis. The proposed method combines the advantages of a new sequential supervised canonical correlation analysis (CCA), introduced in this paper, with the theory of rough hypercuboid approach. While the proposed CCA efficiently combines the local textural information of HEp-2 cells, derived from various scales of rotation-invariant local binary patterns, the relevant and significant features of HEp-2 cell for staining pattern recognition are extracted using rough hypercuboid approach. Finally, the support vector machine, with radial basis function kernel, is used to recognize one of the known staining patterns present in IIF images. The effectiveness of the proposed staining pattern recognition method, along with a comparison with related approaches, is demonstrated on MIVIA, SNP and ICPR HEp-2 cell image databases. An important finding is that the proposed method performs significantly better than state-of-the art methods, on three HEp-2 cell image databases with respect to both classification accuracy and F1 score.

The recently established wireless capsule endoscopy promises itself to be a great advancement in the field of medical diagnosis. The amount of data generated due to a single inspection of gastrointestinal tract is huge, and the transmission demands more cost and power. Due to the inherent nature of the capsule, a low complexity and low-power design are highly recommended with the ability to reduce the data. This paper explains the design of a compression module which satisfies the above requirements. The compression algorithm is developed around some features of the endoscopic images with its suitability for hardware implementation. The preprocessing operations followed by two-stage encoder using differential coder and Golomb–Rice coder are implemented on the experimental dataset. The core module in the preprocessing stage is the chroma subsampling which can considerably reduce the data to be transmitted. Several patterns for subsampling have been analyzed in this paper to evaluate the performance of the proposed system. With the help of a suitably designed up-sampler at the receiver, a near-lossless compression with good reconstructed image quality can be achieved. The 22:1:2 pattern yields better results and could bring about an average peak signal-to-noise ratio of 37 dB with compression rate of 70%. The results show that the proposed chroma subsampling patterns performs better than the competing methods towards wireless capsule endoscopy compression.

General sparse matrix-sparse matrix (SpGEMM) multiplication is one of the basic kernels in a great many applications. Several works focus on various optimizations for SpGEMM. To fully exploit the powerful computing capability of the Sunway TaihuLight supercomputer for SpGEMM, this paper designs the partitioning method and parallelization of CSR-based SpGEMM to make it well match to the Sunway architecture. In addition, this paper optimizes the partitioning method based on the distribution of the floating-point calculations of the CSR-based SpGEMM to achieve the load balance and performance improvement on the Sunway. We, respectively, analyze the performance, including the memory footprint and the execution time, of the parallel CSR-based SpGEMM and the optimized CSR-based SpGEMM on the Sunway. The experimental results show that the optimized CSR-based SpGEMM outperforms over the parallel CSR-based SpGEMM and has good scalability on the Sunway.

With the rapid development of the global economy, people’s living standards have gradually improved, and the way of transportation has undergone earth-shaking changes. This has created a series of social and environmental issues while greatly facilitating people’s lives and work. Excessive numbers of cars have caused urban traffic problems such as traffic pollution, traffic jams and traffic accidents. The strong economic strength has made the automobile industry appear in a booming stage and made it enter the automobile society earlier. The license plate is a vehicle-passed ID card, so the detection and identification of the license plate has become one of the important research directions in today’s society. Although the resolution of today’s surveillance equipment is getting higher, there are still quite a few monitors with low resolution. Conventional license plate detection and recognition algorithms are challenged in low-resolution video processing. In this paper, we introduce the theory and methodology of GAN and transfer learning, which are applied to deal with the license plate image recognition under several complex environments. The experimental results show that the method adopted in this paper has higher recognition rate and robustness.

For pedestrian detection, many deep learning approaches have shown effectiveness, but they are not accurate enough for the positioning of obstructed pedestrians. A novel segmentation and context network (SCN) structure is proposed that combines the segmentation and context information for improving the accuracy of bounding box regression for pedestrian detection. The SCN model contains the segmentation sub-model and the context sub-model. For separating the pedestrian instance from the background and solving the pedestrian occlusion problem, this paper uses the segmentation sub-model for extracting pedestrian segmentation information to generate more accurate pedestrian regions. Considering that different pedestrian instances need different context information, this paper uses context regions with different scales to extract context information. For improving the detection performance, this paper uses the hole algorithm in the context sub-model to expand the receptive field of the output feature maps and connect the multi-channel features with the skip layer. Finally, the loss functions of the two sub-models outputs are fused. The experimental results on different datasets validate the effectiveness of our SCN model, and the deeply supervised algorithm has a good trade-off between accuracy and complexity.

A coded caching scheme for efficient storage and access of medical data is discussed in this paper. Being an inspiring technology for improving the availability and sharing of medical data, the coded caching scheme for medical data is not only convenient for the medical staffs to grasp the patients’ health conditions and treatment history in time, but also effective for the patients to obtain improved quality of medical services. In this paper, the coded caching scheme for the combination network which is the model for the medical data sharing is studied. In the combination network, the server communicates with the users via multiple relays, and the relays as well as the users have cache memories. Using the maximum distance separable codes and placement delivery array (PDA) algorithm, the coded placement phase and delivery phase for combination networks are designed. In addition, for combination networks that satisfy the resolvability property, we extend the PDA algorithm to the case where the users’ priority is considered. The proposed scheme greatly reduces the subpacketization level with slightly increasing the transmission rate. The users with higher priority can get the requested content faster. We demonstrate that relays with cache memories can reduce the data transmission in peak time. The gap between the upper bound and lower bound of the transmission rate is gradually decreased with the increase in the users’ memories. The proposed scheme will shorten the latency time of the system and improve system efficiency for storage and access of medical data.

In order to quantify the time-varying dependent structure between the assets and forecast the portfolio risk accurately, the difference in the preferences for asset risk is taken into consideration in this paper. It is assumed that the new interest rate of asset return sequence is subject to the standard t distribution. A kind of artificial neural network economic forecasting model is put forward. The two-step state transition diagram estimation method for the economic forecasting is deduced, and the forecasting method for the profile risk is obtained. Finally, Shanghai Securities Composite Index and Standard & Poor’s 500 Index are selected to verify the feasibility and superiority of the model and method put forward in this paper. At the same time, the model can accurately quantify the time-varying dependent structural characteristics of the two indices after the subprime mortgage crisis.

This paper mainly analyzes the changing trend of corn price and the factors that affect the price of corn. Using the data and regression analysis, the univariate nonlinear and multivariate linear regression models are established to predict the corn price, respectively. First, this paper establishes a univariate nonlinear regression model with time as the independent variable, and corn price is used as the dependent variable through the analysis of the trend of big data related to Chinese corn price from 2005 to 2016 by MATLAB, which is the computer-based analysis and processing method. The variation of the maize price with time was fitted. To a certain extent, the price trend of corn is predicted. However, the estimated price of corn in 2017 with this model will deviate from the actual value. According to the changes of related policies in our country, we analyzed the deviation of the original model, and the relationship between supply and demand is the main underlying factor that affects the price of corn. This paper selects maize-related big data from 2005 to 2016, we set its production consumption, import and export volume as independent variables, and we still use maize price as the dependent variable to establish a multiple linear regression model. At this stage, the time series analysis of the independent variable has obtained the forecast value of each independent variable in 2017, and then the model is used to predict the corn in 2017 more accurately.

In oil seismic exploration, automatic picking of first arrivals is very important but difficult for seismic data with low signal-to-noise ratios. In the paper, we proposed a new method based on mutual information in information theory. Mutual information between signals and noises is zero theoretically; thus, random noises have fewer effects on first arrivals pickup with the criterion of mutual information. The paper presents a detailed description of the method based on mutual information. Tests and verification with simulation data and field measurements are carried out. When compared with other well-known methods, the first break picking accuracy and efficiency of the proposed algorithm are verified.

This paper proposes the neural network (NN) predictive controller that combines the advantages of NN and predictive control for the automatic voltage regulator (AVR). The NN predictive controller is suggested as a new intelligence controller rather than the conventional controllers for the AVR. This is the first application of the NN predictive controller for AVR. There are five parameters of the NN predictive controller which need a proper tuning to get a good performance by using the NN predictive controller. In recent papers, the parameters of NN predictive controller are typically set by trial and error or by the designer’s expertise. The imperialist competitive algorithm (ICA) is introduced in this paper as a new artificial intelligence technique instead of the trial-and-error or the designer’s expertise methods to get the optimal parameters of NN predictive controller in order to overcome the deviations of the voltage. The performance of the designed NN predictive controller based on the ICA is compared with the designed NN predictive controller based on the genetic algorithm and the conventional proportional–integral–derivative controller based on Ziegler–Nichols technique. The comparison emphasizes the superiority of the suggested NN predictive controller based on the ICA.

Automatic detection of falls is extremely important, especially in the remote monitoring of elderly people, and will become more and more critical in the future, given the constant increase in the number of older adults. Within this framework, this paper deals with the task of evaluating several artificial intelligence techniques to automatically distinguish between different activities of daily living (ADLs) and different types of falls. To do this, UniMiB SHAR, a publicly available data set containing instances of nine different ADLs and of eight kinds of falls, is considered. We take into account five different classes of classification algorithms, namely tree-based, discriminant-based, support vector machines, K-nearest neighbors, and ensemble mechanisms, and we consider several representatives for each of these classes. These are all the classes contained in the Classification Learner app contained in MATLAB, which serves as the computational basis for our experiments. As a result, we apply 22 different classification algorithms coming from artificial intelligence under a fivefold cross-validation learning strategy, with the aim to individuate which the most suitable is for this data set. The numerical results show that the algorithm with the highest classification accuracy is the ensemble based on subspace as the ensemble method and on KNN as learner type. This shows an accuracy equal to 86.0%. Its results are better than those in the other papers in the literature that face this specific 17-class problem.

In view of the problems existing in the evaluation index selection and whether the evaluation method is too simple to evaluate the coordinated development of regional economy. Based on the comprehensive analysis of the definition of the coordinated development of regional economy, this paper discusses the definition of the scientific connotation of the coordinated development of regional economy and puts forward three evaluation indexes for coordinated development of regional economy (the degree of regional economic integration, the degree of regional economic development gap, and the speed of regional economic development). In view of the diversity of evaluation factors and the fuzziness of the relationship between evaluation indexes, this paper uses the fuzzy classification statistical model to evaluate the coordinated development of regional economy and established a regional economic development coordination management system based on fuzzy hierarchical statistical model to evaluate the degree of regional economic coordinated development. Finally, the eastern region and the central and western regions are taken as the object of study. The analysis results show that the conclusion drawn from the model is basically consistent with the actual situation in China. This model can effectively manage the coordinated development of regional economy in China.

At present, there are many methods to realize the matching of specified images with features, and the basic components include image feature point detection, feature description, and image matching. Based on this background, this article has done different research and exploration around these three aspects. The image feature point detection method is firstly studied, which commonly include image edge information-based feature detection method, corner information-based detection method, and various interest operators. However, all of the traditional detection methods are involved in problems of large computation burden and time consumption. In order to solve this problem, a feature detection method based on image grayscale information-FAST operator is used in this paper, which is combined with decision tree theory to effectively improve the speed of extracting image feature points. Then, the feature point description method BRIEF operator is studied, which is a local expression of detected image feature points based on descriptors. Since the descriptor does not have rotation invariance, the detection operator is endowed by a direction that is proposed in this paper, and then the local feature description is conducted on the feature descriptor to generate a binary string array containing direction information. Finally, the feature matching machine learning method is analyzed, and the nearest search method is used to find the nearest feature point pair in Euclidean distance, of which the calculation burden is small. The simulation results show that the proposed nearest neighbor search and matching machine learning algorithm has higher matching accuracy and faster calculation speed compared with the classical feature matching algorithm, which has great advantages in processing a large number of array images captured by the light field camera.

This paper introduces a computer-aided classification (CAC) system for the severity analysis of mitral regurgitation (MR) utilizing multiresolution local binary pattern variants texture features. Initially, the Gaussian pyramid has been used as a multiresolution technique. Subsequently, seven variants of the local binary pattern (LBP) have been employed to extract the features. At last, support vector machine and random forest classifiers are used for classification. The performances of conventional LBP variants and proposed features have been evaluated on MR image database in three classes, i.e., mild, moderate, and severe, in three different views. The Gaussian pyramid-based center-symmetric local binary pattern performed well in all three views. The achieved classification accuracies are 95.66 ± 0.98% in the apical 2 chamber, 94.47 ± 1.91% in the apical 4 chamber and 94.21 ± 1.31% in parasternal long axis views using SVM classifier with the tenfold cross-validation. The outcomes of paper confirm that the performance of the conventional LBP features is enhanced significantly and the proposed CAC system is useful in assisting cardiologists in the severity analysis of MR.

This paper is an addendum to Ref. Neural Computing and Applications (2021) 33:8981–9004 that adds information to clarify certain aspects of the models used adding information and some citations. In this paper an approach based on Co-active Neuro-Fuzzy Inference System named CANFIS-ART is proposed to automate data imputation procedure. This model is compared to other state-of-the-art imputation techniques such as its baseline model Artificial Neural Network and statistical methods, using a total of eighteen databases exposed to a perturbation procedure based on the random generation of non-monotone missing values pattern. A comparison of databases imputed by these models using a set of three classifiers were conducted.

This paper introduces a new evolutionary computing method inspired by the seed transmission process of garden balsam. Garden balsam, a beautiful and attractive flower, randomly ejects the seeds within a certain range by virtue of mechanical force originating from cracking of mature seed pods, which is different from natural expansion of most species of plants. The seeds scattered to suitable growth area will have greater reproductive capacity in the next generation, followed by iteration until the most suitable point for growth in a particular space is eventually found. This phenomenon can more intuitively show the process of searching the problem solution space in the optimization problem. The garden balsam optimization algorithm proposed in this paper incorporates two different types of search processes and has a mechanism to maintain population diversity. Through the optimization experiment on 24 constrained optimization problems, the results obtained by using this algorithm are compared with those of some known meta-heuristic search algorithms. The statistical analysis of the experimental results has been implemented by Friedman rank test and Holm–Sidak test. The comparison results verify the effectiveness of the algorithm.

With the vigorous development of the national economy, the pace and scale of urban construction have been unfolded at an unprecedented speed. A large number of construction projects have made the urban engineering geological exploration activities reach a considerable scale in depth and breadth. The survey results of these projects are very valuable information resources, which not only played an important role in urban planning and construction at that time, but also had high reuse value. Based on BP neural network theory, this paper uses engineering geological database as the research and development platform. Based on the theory of BP neural network and the engineering geological database as the research and development platform, this paper establishes the prediction of geotechnical parameters based on the analysis of the characteristics of geotechnical materials and the distribution of geotechnical sediments and geotechnical parameters. Based on the survey data and specific engineering information, the prediction model of the project was established, and the distribution of the stratum and the relevant geotechnical parameters were predicted. Based on the study of geotechnical properties and BP neural network, a new parameter prediction model is established. Taking the engineering geological database as the platform, using the programming language such as MATLAB, the preliminary research and construction of this prediction system were carried out. The results show that the generalization ability of the prediction model meets the requirements.

Aiming at the problems of long measurement period of online measurement parameters and untimely feedback of IoT technology based on wireless sensor network, this paper designs a multi-parameter online measurement method based on BP neural network algorithm. The collection, analysis, processing and display of parameters are completed through the sensing layer, the network transmission layer and the integrated application layer. The BP neural network algorithm is added to the integrated application layer to optimize the real-time acquisition parameters to shorten the parameter measurement time and accurate prediction. That is, based on the collection of environmental information, by training and learning the BP model with known historical data, it is possible to predict the environmental value at a later time. The known three-layer forward propagation (BP) neural network has the property of approximating the nonlinear curve, and it can achieve good results by predicting the temperature trend. The experimental results show that the system has better ability to monitor and predict the temperature trend. The algorithm simulation experiment shows that the online measurement system based on BP neural network algorithm proposed in this paper can speed up the data collection time, accurately predict the trend of environmental parameters and provide timely warning for potential safety hazards.

In this paper, we use the iterative learning control algorithm to deal with generalized distributed parameter system with parabolic type which described by generalized partial differential equation. Because of the particularity of the generalized system, we usually need generalized value decomposition, but the algorithm we proposed in this paper does not need to consider the impulsive solution of the generalized system, so it can simplify the calculation process. A novel generalized theoretical result is presented by using the PD-type learning law under some assumptions. The convergence conditions of algorithm are established. By the basic generalized theory, matrix theory and mapping principle, the paper gives rigorous convergence proof of the algorithm to ensure the tracking error is convergent in L2 norm. Finally, we use an example to verify the validity of the new proposed algorithm. Through this paper, we can do further study and discuss the iterative learning algorithm for generalized distributer parameter system in the future.

This paper presents a novel, fully automatic method to grayscale images colorization using a scene guided neural network. In our proposed method, given a training set of both grayscale images and their corresponding color images, we first extract features of each grayscale pixel. These features, together with the corresponding RGB values of that pixel are input to train a colorization neural network for each given scene. To improve the performance of colorization, in both speed and results, we further classify the input and training images into different scene classes. We adopt a linear image classification method to generate a scene guided codebook and use it to determine the scene class of the input image. The preliminary colorization result is then generated by the corresponding trained neural network of the scene class of the input image. Finally, an image guided filter is used to refine colorized images. Inspired by the recent success in deep learning techniques which provide stabilizing modeling of large-scale medical image data, the proposed paper formulating the enhancement and colorization problem, so that colorization techniques can be directly used to ensure medical images with high quality. The experimental results on a broad range of images demonstrate that our method has better colorization performance as compared to that of the state-of-the-art algorithms.

In the field of biometrics, ear recognition is a niche idea of recognition for human authentication, which has several merits compared to the other biometric recognitions like face and finger print. The contour of an ear is distinctive for each person, which is the main reason for choosing this recognition technique. Only in very few studies, the ear recognition algorithms were presented. There still remains a large space for research in the field of ear biometrics. All recent papers have implemented ear recognition algorithms using 2-D ear images. The ear recognition algorithms should be efficient in order to provide accurate results, owing to issues like multiple poses and directional related. This paper proposes a novel method for segmentation based on adaptive approach Runge–Kutta (AARK) to recognize ear images. AARK threshold segmentation technique is used for finding the threshold value to determine the region to be segmented. The utilization of AARK’s numerical methods in computing the threshold value for ear recognition process improves the result accuracy. Firstly, preprocessing has been carried out for the dataset. The following steps are carried out sequentially: ring projection, information normalization, morphological operation, AARK segmentation, feature extraction of DWT and finally ANFIS classifier are used. Among the various steps mentioned, ring projection converts the two dimensions into single dimensions. The self-adaptive discrete wavelet transform is used to extract features from the segmented region. Then the ANFIS classifier is used to recognize the ear region from the image by taking the features form the test image and the training images. The proposed method obtained 72% improvement in PSNR and accuracy is improved to 63.3%. Moreover, the speed and space occupation of the self-adaptive DWT technique and the conventional DWT technique are measured by implementing the methods in FPGA Spartan 6 device. Comparing with the implementation of conventional DWT, the area is reduced to 361 from 7021 while implementing the proposed self-adaptive DWT method.

This paper focuses on the core problem of database resource integration mode. The theme of this paper is the construction and implementation of database resource integration model, and some legal and commercial constraints related to it, so the content of the research involves the integration of technology, integration system, the integration of the status analysis and other aspects. The basic concept, classification, structure system and basic principles of Oracle RAC are discussed. Combined with business needs, a solution based on Oracle RAC is proposed. We implemented the planning and deployment of shared storage, file storage and network setup based on Oracle RAC database. The automatic failover without user intervention is realized, which greatly improves the reliability of the database, and has a certain guiding significance for the application and construction of the enterprise database. First, the research status of database resources integration at home and abroad is expounded through literature analysis. Secondly, the model of database resource integration is put forward, and its realization is discussed. Thirdly, the status of database resource integration is analyzed. Finally, the problems to be solved are put forward in the end. Through the research of this topic, it provides some reference and basis for the database resource integration model. At the same time, it also makes a paving for the maximum elimination of information isolated island and promoting the sharing of database resources.

In this paper, a new approach of automatic building for domain ontology based on machine learning algorithm is proposed, and by which the large-scale e-Gov ontology is built automatically. The advent of the knowledge graph era puts forward higher requirements for semantic search and analysis. Since traditional manual ontology construction requires the participation of domain experts in large-scale ontology construction, which will take time and considerable resources, and the ontology scale is also limited. The approach proposed in this paper not only makes up for the shortage of thesaurus description of the semantic relation between terms, but also takes advantage of the massive online encyclopedia knowledge and typical similarity algorithm in machine learning to fill the domain ontology automatically, so that the advantages of the two different knowledge sources are fully utilized and the system as a whole is gained. Ultimately, this may provide the foundation and support for the construction of knowledge graph and the semantic-oriented applications.

Based on the hardware and sensors of image acquisition, the noise in the image has been easily generated. In this paper, an improved method of image decompression has proposed the shortcoming of the above-mentioned hardware algorithm. The traditional filter desiccation algorithm can only remove one or two specific noises, and it is not effective for other types. We combine some excellent neural network models. In this paper, an image mixing noise removal algorithm based on convolution nerve has been mentioned. Aiming at realizing the super-resolution of the image, the deconvolution layer can be used only to enlarge the image. The magnification factor is the step of deconvolution. This paper aims to eliminate the interference of the image noise. The effect of magnification on the deconvolution layer is impossible. The results of experimental test show that the algorithm achieves a good noise removal effect and is suitable for various mixed noise images. The algorithm used in this paper improves the subjective visual effect and objective evaluation index.

Hand belongs to non-rigid objects and is rich in variety, making gesture recognition more difficult. The essence of dynamic gesture recognition is the classification and recognition of single-frame still images. Therefore, this paper mainly focuses on static gesture recognition. At present, there are some problems in gesture recognition, such as accuracy, real-time or poor robustness. To solve the above problems, in this paper, the Kinect sensor is used to obtain the color and depth gesture samples, and the gesture samples are processed. On this basis, a jointly network of CNN and RBM is proposed for gesture recognition. It mainly uses superposed network of multiple RBMs to carry out unsupervised feature extraction and combined with supervised feature extraction of CNN. Finally, these two features are combined to classify them. The simulation results show that the proposed jointly network has a better performance in identifying simple background gesture samples and the recognition capability of gesture samples in complex background needs to be improved.

Deep convolutional neural networks (CNNs) have recently shown very high accuracy in a wide range of cognitive tasks, and due to this, they have received significant interest from the researchers. Given the high computational demands of CNNs, custom hardware accelerators are vital for boosting their performance. The high energy efficiency, computing capabilities and reconfigurability of FPGA make it a promising platform for hardware acceleration of CNNs. In this paper, we present a survey of techniques for implementing and optimizing CNN algorithms on FPGA. We organize the works in several categories to bring out their similarities and differences. This paper is expected to be useful for researchers in the area of artificial intelligence, hardware architecture and system design.

The maximum independent set algorithm for large-scale data semi-existing data is studied and the solving method of the largest independent set problem in large-map data is mainly analyzed in this paper. The specific research contents are mainly divided into semi-external map algorithm based on Greedy heuristic strategy, semi-external map algorithm based on swap and design, and implementation of semi-external graph algorithm processing function library. Experiments on a large number of real and artificially generated data sets show that the algorithm in this paper is very efficient both in time and in space. The largest independent set obtained by the algorithm can reach more than 96% of its theoretical upper bound for most of the data.