Search

An improved local non-negative matrix factorizetion (LNMF) algorithm for palmprint recognition is proposed to solve the problem regarding the high requirements of palmprint image preprocessing. First, the Fisher discriminant criterion is integrated into the non-negative matrix algorithm of the target function, using the gradient descent method. To solve the drawbacks of the traditional algorithm, the LNMF algorithm does not contain class information affecting the recognition rate, and finally, it chooses the appropriate classifier for matching. Experiments on two databases are conducted, and the results show that the proposed method has very high recognition rates on the two databases. In a comparison with other algorithms, the current algorithm achieves a higher recognition effect.

The cuckoo search algorithm (CSA) is a natureinspired metaheuristic algorithm based on the parasitic behaviour of the cuckoo species, and it has received much attention for solving optimization problems. Nevertheless, the CSA still has some limitations when applied to optimization problems, such as easily falling into local optima and converging slowly. In this paper, we propose a modified local search method with an adaptive boundary scaling mechanism using a chaotic map, which calculates the search radius depending on the position of the current population. Thus, the search ability of the traditional CSA is improved for optimization purposes. To further verify the stability and effectiveness of the proposed adaptive boundary scaling algorithm (ACSA), we compare it with the traditional CSA and nonmodified chaos-embedded CSA (CCSA) on 25 test benchmark functions. The experimental and statistical test results show that the ACSA outperforms its competitors, which means that the proposed algorithm can perform better than these other approaches in terms of their search capabilities for optimization problems.

The chaotic and intrinsic complexity of wind speed time series calls for an appropriate model to accurately predict the moving tendency. In this study, we proposed a single dendritic neuron model (S-DNM) using a back-propagation algorithm to accomplish wind speed forecasting. First, based on mutual information method and false nearest neighbor method, the time delay and embedding dimension are calculated. Second, the phase space is reconstructed by time delay and embedding dimension, and the characteristics of wind speed time series are analyzed. Then, the maximum Lyapunov exponent is applied to confirm the chaotic properties of the wind speed time series. Finally, using wind speed data from Sotavento located in Galicia, Spain, the performance of the forecasting method is evaluated for short-term horizons (1 hour ahead). Experimental results show that the proposed S-DNM performed better than the traditional ELMAN model and classic multi-layered perceptron network. Thus, it is concluded that the proposed model is suitable for wind speed forecasting.

This paper reports a wearable and wireless system for Braille recognition, called TouchReader, which utilizes a fast and accessible strategy to facilitate visually impaired people in learning and communication. This finger-worn device is composed of flexible and piezoresistive sensor array with high density (96 sensors in 78.5 mm2), high sensitivity (8.44 kPa−1 in low pressure range) and a template-matching recognition system. It is the first time to achieve area-scanning of Braille character on the strength of tactile sensors and convert text to voice message, exhibiting better practicability, portability and reliability in daily routine with less manual intervention.

Gravitational search algorithm (GSA) has attracted more and more attention in dealing with complex optimization problems. However, it still suffers from some major drawbacks, such as poor exploitation. This paper has proposed a novel self-adaptive chaotic search mechanism embedded GSA (SA-GSA), in which different chaotic maps are utilized to guide the local search in a self-adaptive way. Specifically, instead of defining the search range randomly and arbitrarily, the distance between distinct individuals has been utilized in the current population as the search range. Thus the search range will decrease synchronously according to the convergence speed of the population, which is thought to improve the exploitation ability of GSA effectively. To evaluate the performance of SA-GSA, we compare it with classic GSA and the classic chaotic GSA (CGSA) on 23 benchmark optimization problems. The experimental results and statistic analysis verify that SA-GSA is superior to its competitors in terms of convergence speed and solution accuracy rate.

With recent developments and advances in machine learning methods, traditional time series analysis techniques, e.g., ARMA, ARIMA, ARIMAX, SARIMAX, etc. models are being replaced by deep learning models. The application of Recurrent Neural Network (RNN) and Long-Short Term Memory (LSTM) to the problem of solar irradiance prediction has been shown in the literature to achieve state-of-the-art performance in this domain. For sequence modeling, Temporal Convolutional Networks (TCN) are gaining increasing attention because of the excellent trade-off between performance accuracy and time in training the models. In this paper, an evaluation of these deep learning methods is completed on the application of short-term, one hour ahead solar irradiance forecasting. The results show that TCNs can reduce the resource usage, depending on the data-folding size and model complexity, to require only 5% of the training time with only a loss of 6% RMSE.

Predicting the three-dimensional structure of a protein from its amino acid sequence is an important issue in the field of computational biology and bioinformatics. It remains as an unsolved problem and attract enormous researchers' interests. Different from most conventional methods, we model the protein structure prediction (PSP) problem as a multi-objective optimization problem. A three-objective energy function based on three physical terms is designed to evaluate a protein conformation. A multi-objective evolutionary strategy algorithm coupled with preference information is proposed in this study. The preference information is used in the survival criteria, focusing on the exploration of search process. The experimental results based on five proteins in PDB library demonstrate the effectiveness of proposed method. The analysis of Pareto fronts indicates that the preference information can make solutions diverse in genotypic space. Thus, the proposed method gives a new perspective for solving PSP problems.

Recently, more neuroscience researches focus on the role of dendritic structure during the neural computation. Inspired by the specified topologies of numerous dendritic trees, we proposed a single neural model with a particular dendritic structure. The dendrites are composed of several branches, and these branches correspond to three distributions in coordinate, which are used to classify the training data as required. Genetic algorithm is used as the training algorithm. Experimental results based on two benchmark classification problems verify the effectiveness of the proposed method, and the distributions of trained dendritic structures are also presented.

IEEE Std 1641 defines a method of accurately defining signals as used in test procedures and requirements. It defines many common signals in the basic signal component (BSC) and test signal framework (TSF) layers. New signals can be described by combining existing signals, and there is no limit on the expansibility of signals.

The risk comes from nature and factitious attack for POL transportation. According to design united transportation network, the paper takes into consideration of mini-risk route zone risk equity and at the same time. This can decrease the risk for POL in transportation, effectively and increase the transportation safety.

China is experiencing rapid urbanization stage of development, and the environmental problems are increasingly severe. For the healthy development of the city, the planning and construction of country parks is being more and more attention. The Sai Kung West Country Park is a typical and mature case of the country park system in HongKong. Based on field survey, the authors try to interpretation the main content and regional characteristics in its planning and design, and also try to sum up the useful inspiration for the Country Park design and construction in China.

With the national implementation of the western development strategy in China, Southwest campus construction have been more and more attention. The main difference between the southwest campus and traditional campus is that the former more emphasizes on the efficient design and construction of land suitability, pursuit of harmony between man and nature, try to achieve ecological balance and sustainable development. Based on the design practice in new campus planning of Zunyi Medical College, the article explored the regional design strategy by taking full account of topography and climate characteristics.

An efficient digital notch filter for 50Hz noise eliminated from low frequency photoacoustic gas sensing signal is presented in this paper by a Matlab program. The digital notch filter is consisted of the fast Fourier transform, a second order infinite impulse response notch filter, inverse fast Fourier transform, real and imaginary parts separating, phase shift for imaginary part, and pure noise eliminating from real and imaginary parts. A smooth waveform of low frequency photoacoustic signal can be achieved through the 50Hz digital notch filter and can be used as the analysis data to calculate the target gas concentration. The detectable sensitivity of the photoacoustic gas sensing system with the 50Hz digital notch filter can be obtained as 100 parts-per-billion acetylene when about 0.2 percent water vapor exists and interferes for the low concentration acetylene at 1530.37nm.