MELM-GRBF: A modified version of the extreme learning machine for generalized radial basis function neural networks

In this paper, we propose a methodology for training a new model of artificial neural network called the generalized radial basis function (GRBF) neural network. This model is based on generalized Gaussian distribution, which parametrizes the Gaussian distribution by adding a new parameter τ. The generalized radial basis function allows different radial basis functions to be represented by updating the new parameter τ. For example, when GRBF takes a value of τ=2, it represents the standard Gaussian radial basis function. The model parameters are optimized through a modified version of the extreme learning machine (ELM) algorithm. In the methodology proposed (MELM-GRBF), the centers of each GRBF were taken randomly from the patterns of the training set and the radius and τ values were determined analytically, taking into account that the model must fulfil two constraints: locality and coverage. An thorough experimental study is presented to test its overall performance. Fifteen datasets were considered, including binary and multi-class problems, all of them taken from the UCI repository. The MELM-GRBF was compared to ELM with sigmoidal, hard-limit, triangular basis and radial basis functions in the hidden layer and to the ELM-RBF methodology proposed by Huang et al. (2004) [1]. The MELM-GRBF obtained better results in accuracy than the corresponding sigmoidal, hard-limit, triangular basis and radial basis functions for almost all datasets, producing the highest mean accuracy rank when compared with these other basis functions for all datasets.     

Gaussian kernel optimization: Complex problem and a simple solution

The Gaussian kernel function implicitly defines the feature space of an algorithm and plays an essential role in the application of kernel methods. The parameter of Gaussian kernel function is a scalar that has significant influences on final results. However, until now, it is still unclear how to choose an optimal kernel parameter. In this paper, we propose a novel data-driven method to optimize the Gaussian kernel parameter, which only depends on the original dataset distribution and yields a simple solution to this complex problem. The proposed method is task irrelevant and can be used in any Gaussian kernel-based approach, including supervised and unsupervised machine learning. Simulation experiments demonstrate the efficacy of the obtained results. A user-friendly online calculator is implemented at: www.csbio.sjtu.edu.cn/bioinf/kernel/ for public use.     

Effect of micro-electrode geometry on NO2 gas-sensing characteristics of one-dimensional tin dioxide nanostructure microsensors

Electrodes with micro-gaps are fabricated by using dc-sputtering and FIB techniques. SnO₂ nanowires are deposited on the micro-gap (1-30 μm) by suspension dropping method to fabricate a micro-gas sensor. The sensing ability of various SnO₂ micro-gap sensors is measured. A comparison between sensors reveals that the short-gap electrode has numerous advantages in terms of reliability, high sensitivity and detection of low concentrations of NO₂, while the large-gap electrode is relatively sensitive for high concentrations. Conductance measurements are carried out at different surface temperatures and NO₂ concentrations in order to investigate the effects that the gap size has on the overall sensor conductance. The results suggest that the interface between the electrode and sensitive layer has a very important role for the sensing mechanism of tin dioxide gas sensors.     

改进的基于内容的文件类型识别算法

在现有基于内容的文件类型识别算法基础上,针对统计特征提取方面存在的问题,采用定长和变长窗口对文件二进制内容进行划分,提取文件的统计特征,并提出将特征选择应用于文件类型识别,结合特征的广度和稳定度设计出一种特征选择评估函数选择标志特征,从而建立文件类型模型,以此为标准识别文件类型.该算法不依靠特定文件类型的结构和关键标识...     

A computer-enabled solution procedure for food wholesalers’ distribution decision in cities with a circular transportation infrastructure

This paper addresses the food distribution decision from a wholesaler's perspective, as the wholesalers are often faced with thorny issues such as when, where, how to deliver products at the lowest costs to best satisfy the retailers’ changing needs. This research relies on a distribution problem encountered at Northern Grocery Company in Beijing as an example to develop a two-stage solution procedure and subsequent computer-enabled programs for identifying efficient vehicle routing alternatives. The solution process, which integrates heuristic search algorithms and an integer programming model, can be applied to a wide range of distribution problems with specified route durations and a circular transportation network structure as exemplified by City of Beijing.     

Effects of differences in office chair controls, seat and backrest angle design in relation to tasks

In this study the influence of chair characteristics on comfort, discomfort, adjustment time and seat interface pressure is investigated during VDU and non-VDU tasks: The two investigated office chairs, both designed according to European and Dutch standards are different regarding: 1) seat cushioning and shape, 2) backrest angle and 3) controls. Thirty subjects in total, both male and female, participated in two experiments: twenty in the first and ten in the second.Significant differences are found for ease of adjustment and adjustment time of controls, independent of the tasks. Related to tasks, a significant difference was found for the backrest range of motion. For non-VDU tasks a larger range of backrest motion was preferred by 70% of the subjects. The chair design differences were most clear for comfort and adjustment time of controls, followed by comfort of backrest angle. No differences are found between seat pan comfort and discomfort, first impressions and peak interface pressure.     

A novel user identification scheme with key distribution preserving user anonymity for distributed computer networks

Recently, Yang et al. proposed an efficient user identification scheme with key distribution, in which it is possible for the user to anonymously log into a system and establish a secret key shared with the system. Mangipudi and Katti later demonstrated a Deniable-of-Service (DoS) attack on the Yang et al. scheme and then proposed an improvement to withstand such an attack. However, this paper demonstrates an identity disclosure attack to show that neither schemes’ claimed user anonymity requirement can be achieved. We further propose a novel user identification scheme with key distribution preserving user anonymity for distributed computer networks. The proposed scheme not only withstands the attacks mentioned above, but also achieves the following: (i) user anonymity, (ii) key distribution, (iii) mutual authentication, and (iv) key confirmation. The performance of our scheme is of greater efficiency than that of previously proposed schemes in terms of communication costs and computational complexities.     

Discrete data clustering using finite mixture models

Finite mixture models have been applied for different computer vision, image processing and pattern recognition tasks. The majority of the work done concerning finite mixture models has focused on mixtures for continuous data. However, many applications involve and generate discrete data for which discrete mixtures are better suited. In this paper, we investigate the problem of discrete data modeling using finite mixture models. We propose a novel, well motivated mixture that we call the multinomial generalized Dirichlet mixture. The novel model is compared with other discrete mixtures. We designed experiments involving spatial color image databases modeling and summarization, and text classification to show the robustness, flexibility and merits of our approach.     

Estimating the altitude of aerosol plumes over the ocean from reflectance ratio measurements in the O2 A-band

A methodology is proposed to infer the altitude of aerosol plumes over the ocean from reflectance ratio measurements in the O₂ absorption A-band (759 to 770 nm). The reflectance ratio is defined as the ratio of the reflectance in a first spectral band, strongly attenuated by O₂ absorption, and the reflectance in a second spectral band, minimally attenuated. For a given surface reflectance, simple relations are established between the reflectance ratio and the altitude of an aerosol layer, as a function of atmospheric conditions and the geometry of observation. The expected accuracy for various aerosol loadings and models is first quantified using an accurate, high spectral resolution, radiative transfer model that fully accounts for interactions between scattering and absorption. The method is developed for POLDER and MERIS, satellite sensors with adequate spectral characteristics. The simulations show that the method is only accurate over dark surfaces when aerosol optical thickness at 765 nm is relatively large (> 0.3). In this case, the expected accuracy is on the order of ± 0.5 km or ± 0.2 km for POLDER or MERIS respectively. More accurate estimates are obtained with MERIS, since in this case the spectral reflectance ratio is more sensitive to aerosol altitude. However, a precise spectral calibration is needed for MERIS. The methodology is applied to MERIS and POLDER imagery acquired over marine surfaces. The estimated aerosol altitude is compared with in situ lidar profiles of backscattering coefficient measured during the AOPEX-2004 experiment for MERIS, or obtained with the space-borne lidar CALIOP for POLDER. The retrieved altitudes agree with lidar measurements in a manner consistent with theory. These comparisons demonstrate the potential of the differential absorption methodology for obtaining information on aerosol altitude over dark surfaces.