Accelerating the kernel-method-based feature extraction procedure from the viewpoint of numerical approximation

The kernel method suffers from the following problem: the computational efficiency of the feature extraction procedure is inversely proportional to the size of the training sample set. In this paper, from a novel viewpoint, we propose a very simple and mathematically tractable method to produce the computationally efficient kernel-method-based feature extraction procedure. We first address the issue that how to make the feature extraction result of the reformulated kernel method well approximate that of the naïve kernel method. We identify these training samples that statistically contribute much to the feature extraction results and exploit them to reformulate the kernel method to produce the computationally efficient kernel-method-based feature extraction procedure. Indeed, the proposed method has the following basic idea: when one training sample has little effect on the feature extraction result and statistically has the high correlation with regard to all the training samples, the feature extraction term associated with this training sample can be removed from the feature extraction procedure. The proposed method has the following advantages: First, it proposes, for the first time, to improve the kernel method through formal and reasonable evaluation on the feature extraction term. Second, the proposed method improves the kernel method at a low extra cost and thus has a much more computationally efficient training phase than most of the previous improvements to the kernel method. The experimental comparison shows that the proposed method performs well in classification problems. This paper also intuitively shows the geometrical relation between the identified training samples and other training samples.     

Effects of the number of hidden nodes used in a structured-based neural network on the reliability of image classification

A structured-based neural network (NN) with backpropagation through structure (BPTS) algorithm is conducted for image classification in organizing a large image database, which is a challenging problem under investigation. Many factors can affect the results of image classification. One of the most important factors is the architecture of a NN, which consists of input layer, hidden layer and output layer. In this study, only the numbers of nodes in hidden layer (hidden nodes) of a NN are considered. Other factors are kept unchanged. Two groups of experiments including 2,940 images in each group are used for the analysis. The assessment of the effects for the first group is carried out with features described by image intensities, and, the second group uses features described by wavelet coefficients. Experimental results demonstrate that the effects of the numbers of hidden nodes on the reliability of classification are significant and non-linear. When the number of hidden nodes is 17, the classification rate on training set is up to 95%, and arrives at 90% on the testing set. The results indicate that 17 is an appropriate choice for the number of hidden nodes for the image classification when a structured-based NN with BPTS algorithm is applied.     

Generalized Possibilistic Fuzzy C-Means with novel cluster validity indices for clustering noisy data

A generalized form of Possibilistic Fuzzy C-Means (PFCM) algorithm (GPFCM) is presented for clustering noisy data. A function of distance is used instead of the distance itself to damp noise contributions. It is shown that when the data are highly noisy, GPFCM finds accurate cluster centers but FCM (Fuzzy C-Means), PCM (Possibilistic C-Means), and PFCM algorithms fail. FCM, PCM, and PFCM yield inaccurate cluster centers when clusters are not of the same size or covariance norm is used, whereas GPFCM performs well for both of the cases even when the data are noisy. It is shown that generalized forms of FCM and PCM (GFCM and GPCM) are also more accurate than FCM and PCM. A measure is defined to evaluate performance of the clustering algorithms. It shows that average error of GPFCM and its simplified forms are about 80% smaller than those of FCM, PCM, and PFCM. However, GPFCM demands higher computational costs due to nonlinear updating equations. Three cluster validity indices are introduced to determine number of clusters in clean and noisy datasets. One of them considers compactness of the clusters; the other considers separation of the clusters, and the third one considers both separation and compactness. Performance of these indices is confirmed to be satisfactory using various examples of noisy datasets.     

Vessel segmentation and width estimation in retinal images using multiscale production of matched filter responses

Automated segmentation of blood vessels in retinal images can help ophthalmologists screen larger populations for vessel abnormalities. However, automated vessel extraction is difficult due to the fact that the width of retinal vessels can vary from very large to very small, and that the local contrast of vessels is unstable. Further, the small vessels are overwhelmed by Gaussian-like noises. Therefore the accurate segmentation and width estimation of small vessels are very challenging. In this paper, we propose a simple and efficient multiscale vessel extraction scheme by multiplying the responses of matched filters at three scales. Since the vessel structures will have relatively strong responses to the matched filters at different scales but the background noises will not, scale production could further enhance vessels while suppressing noise. After appropriate selection of scale parameters and appropriate normalization of filter responses, the filter responses are then extracted and fused in the scale production domain. The experimental results demonstrate that the proposed method works well for accurately segmenting vessels with good width estimation.     

Behaviour of concrete-filled steel tubular columns incorporating fly ash

The subject of this work is to investigate the effect of fly ash on the strength of concrete filled steel tubular columns from 28 to 365 days. A contrast study was carried out on concrete filled steel tubular columns incorporating 10–40 wt% fly ash, and for control Portland cement concrete filled steel tubular columns. The effect of pre-coating the inner surface of steel tubes with a thin layer of fly ash was also studied. Assessments of the concrete mixes were based on the compressive strength and the bond strength. The results show that a lower replacement with fly ash can improve both bond strength and compressive strength, while a higher replacement with fly ash requires a relatively longer time to achieve similar beneficial effects. Pre-coating the inner surface of steel tubes with a thin layer of fly ash can notably improve the bond strength. The microstructure of the interface between concrete and steel tube was also studied by using scanning electron microscopy analyzer.     

选矿厂的DH+网络自动监控与管理系统

结合工程实例,采用DH 网络技术、PLC控制技术及工业自动化组态软件技术实现了选矿厂自动监控和管理系统。系统实现了给矿自动控制、磨矿生产过程自动控制和对整个生产的自动监控与管理,对提高生产效率和产品质量起到了重要作用。指出DH 网络系统的合理配置可优化系统结构、降低成本,防干扰措施的合理实施是保证系统稳定运行的重要环节。     

Improved robust H2 and H∞ filtering for uncertain discrete-time systems

This paper is concerned with the robust H₂ and H_∞ filtering problems for linear discrete-time systems with polytopic parameter uncertainty. We aim to derive a less-conservative design than existing linear matrix inequality (LMI) based sufficient conditions. It is shown that a more efficient evaluation of robust H₂ or H_∞ performance can be obtained by a matrix inequality condition which contains additional free parameters as compared to existing characterizations. When applying this new matrix inequality condition to the robust filter design, these parameters provide extra degrees of freedom in optimizing the guaranteed H₂ or H_∞ performance and lead to a less-conservative design.     

ε constrained differential evolution with pre-estimated comparison using gradient-based approximation for constrained optimization problems

Many real-world problems can be categorized as constrained optimization problems. So, designing effective algorithms for constrained optimization problems become more and more important. In designing algorithms, how to guide the individuals moving more efficiently towards the feasible region is one of the most important aspects on finding the optimum of constrained optimization problems. In this paper, we propose an improved ε constrained differential evolution, which combines with pre-estimated comparison gradient based approximation. The proposed algorithm uses gradient matrix to determine whether the trail vector generated by differential evolution algorithm is worth using the fitness function to evaluate it or not. Pre-estimated comparison gradient based approximation is used as a detector to find the promising offspring and in this way can we guide the individuals moving towards the feasible region. The proposed method is tested both on twenty-four benchmark functions and four well-known engineering optimization problems. Experimental results show that the proposed algorithm is highly competitive in comparing with other state-of-the-art algorithms. The proposed algorithm offers higher accuracy in engineering optimization problems for constrained optimization problems.     

One-step production of biodiesel through simultaneous esterification and transesterification from highly acidic unrefined feedstock over efficient and recyclable ZnO nanostar catalyst

Zinc oxide (ZnO) nanostar synthesized by simple and up-scalable microwave-assisted surfactant free hydrolysis method was applied as catalyst for biodiesel synthesis through one-step simultaneous esterification and transesterification from high free fatty acid (FFA) contaminated unrefined feedstock. It was found that ZnO nanostar catalyst was reacted with FFA to yield zinc oleate (ZnOl) as intermediate and finally became zinc glycerolate (ZnGly). With the re-deposition of ZnGly back to the ZnO nanostar catalyst at the end of the reaction, the catalyst can be easily recovered and stay active for five cycles. Furthermore, the rate of transesterification is highly promoted by the presence of FFA (6 wt.%) which makes it an efficient catalyst for low grade feedstock like waste cooking oil and crude plant oils.     

Effect of ethanol in the coagulation bath on the structure and performance of PVDF-g-PEGMA/PVDF membrane

This report investigated the effect of ethanol content in the coagulation bath on the surface composition, membrane pore size structure, pure water flux, and permeability of the amphiphilic polymer polyvinylidene fluoride (PVDF)-g-poly (ethylene glycol) methyl ether methacrylate (PEGMA)-modified PVDF membrane. The study found that pore size and their distribution and, as a result, membrane permeability, can also be easily controlled by adjusting ethanol content in the coagulation bath. Membrane water fluxes formed by the coagulation baths with 0, 10, 20, and 30% of ethanol were 1843.65, 2774.61, 4391.88, and 5142.35 L (m⁻²·h⁻¹). When the content of ethanol in the coagulation bath is high, the surface enrichment of PEGMA slightly decreases, and the surface becomes rougher. Thus, the decrease of the hydrophilic functional groups on the surface of the membrane and the increase of the roughness leads to the deterioration of the hydrophilicity of the membranes surface. © 2018 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2019 , 136, 47380.