EEG extended source imaging with structured sparsity and $$L_1$$ -norm residual

Neural Computing and Applications - It is a long-standing challenge to reconstruct the locations and extents of cortical neural activities from electroencephalogram (EEG) recordings, especially...     

Water Bloom Precursor Analysis Based on Two Direction S-Rough Set

In the field of water quality management, it is vital to determine the main precursory anomalies from the precursor of intricate water bloom in the context of a given area. In this paper, a water bloom precursor analysis method, based on two direction singular rough set, was proposed. This approach was produced on the basis of the different sections and pre-water bloom of water bloom precursor anomalies and characteristic of elements transferred in singular rough set. For testing the validity of two direction singular rough set application in water bloom precursor analysis, Xiangxi River, which is one of the typical tributaries of Three Gorges Reservoir in China, was selected as study area. The result showed that compared with other indexes, pH and dissolved oxygen (DO) are the most valuable indicators of water bloom in the precursory anomalies. Furthermore, regarding with water bloom precursory anomalies in Xiangxi River, most of the nutrient loading and biological community are the key indicators. Hence, this method can determine the main precursory anomaly for water bloom in the study area, which provides powerful knowledge support to water quality specialists for them to comprehensively analyze precursory anomaly so as to find out its relationship with occurrence law of water bloom.     

一种基于非残差估计线性表示模型的人脸识别

Dantzig selector利用样本与残差之间的相关向量的L!范数来约束线性表示模型,为了克服传统的线性表示在处理表示残差时依赖于一个对残差特定的似然估计这一缺陷将这样一种处理残差的思路用于人脸识别,提出了一种基于Dantzig selector模型的人脸识别算法,并对Dantzig selector的一种有效的求解方法进行研究。在常用人脸库上实验表明,基于Dantzig selector人脸识别算法在不需要对残差进行估计的情况下也能取得了很好的识别效果。     

Extraction and separation of Ga(Ⅲ) and T1(Ⅰ) or In(Ⅲ) and T1(Ⅰ) from chloride media with CA-12 in sulphonating kerosene

The extraction behavior and mechanism of gallium(Ⅲ) and thallium(Ⅰ) or indium(Ⅲ) and thallium(Ⅰ) with Sec-octylphenoxy acetic acid (CA-12) from hydrochloric acid solution in sulphonating kerosene were studied. Optimum extraction conditions were evaluated from critical study of various experimental parameters and the probable composition of the extracted species is ascertained from lg D-lg C plots. The effects of different ions on the determination of these ions were examined.     

基于轻量级密集残差网络的水下图像增强

深度卷积神经网络是水下图像增强的主要方法之一,但其过高的内存消耗和计算需求阻碍了在实际应用中的部署。为此,提出一种轻量级的密集残差卷积神经网络(dense residual convolutional neural networks, DRCNN)用于水下图像增强。为降低计算成本,DRCNN采用深度可分离卷积提取高级特征;通过密集连接和残差学习促进不同通道之间的信息交互,提高模型表征能力;将输入的退化图像与中间特征图融合,保留图像全局相似性,同时防止模型梯度消失。实验结果证明DRCNN能有效提高水下图像质量,较于现有算法,DRCNN参数量减少了85%,PSNR、SSIM值分别提高了3%、2%,测试速度提高了3%。DRCNN使用更少的参数实现了更好的性能,利于在低资源设备的实时场景中应用。     

TMLNN: triple-valued or multiple-valued logic neural network

We discuss the problem of representing and processing triple-valued or multiple-valued logic knowledge using neural network. A novel neuron model, triple-valued or multiple-valued logic neuron (TMLN), is presented. Each TMLN can represent a triple-valued or multiple-valued logic rule by itself. We will show that there are two TMLNs: TMLN-AND (triple-valued or multiple-valued "logic AND") neuron and TMLN-OR (triple-valued or multiple-valued "logic OR") neuron. Two simplified TMLN models are also presented, and show that a multilayer neural network made up of triple-valued or multiple-valued logic neurons (TMLNN) can implement a triple-valued or multiple-valued logic inference system. The training algorithm for TMLNN is presented and can be shown to converge. In our model, triple-valued or multiple-valued logic rules can be extracted from TMLNN with ease. TMLNN can thus form a base for representing logic knowledge using neural network.     

基于Lotus Notes的分布式信息商函系统的设计与实现

设计并实现了一种基于Ｌｏｔｕｓ Ｎｏｔｅｓ的分布式信息邮政商业信函系统,并研究了系统实现过程中的两个关键技术。     

使用动态矩形窗求最近点对的几何算法

针对几种经典算法的效率问题,提出了一种最近点对求解算法:算法预先找到X、Y坐标轴最大、最小坐标值,将点集按某坐标轴排序,并根据鸽巢原理计算最近点对i间距离的上限,由此上限将整个点集分割成不同的区域,最后在每个区域中通过动态缩小每个点的矩形窗减少了算法的计算次数.通过理论证明和实际验证得出:初始点集有序时算法的时间复杂度为O(n);无序时在一般情况下为O(n),最坏情况下为O(nlogn);算法的时间复杂度和运行时间明显优于经典分治算法和基于底函数的算法.     

含B2O3无氟连铸保护渣物理性能的研究

常规保护渣中配入的助熔剂一般是氟化物和Na2O,但二者都对环境产生严重污染。为此试验研究了含助熔剂B2O3保护渣及其熔点、粘度、结晶性能和表面张力等物理性能。试验表明，当保护渣碱度(CaO/SiO2)等于0．88，成分(％)为37．91CaO-43.09SiO2—5Al2O3—5MgO-2Li2O-7B2O3时的熔点909℃，熔化性温度1160℃，1300℃粘度值0．40Pa．s,1300℃表面张力值0．32N/m，能满足常规钢种和部分特殊钢种对连铸保护渣的使用要求。     

Simultaneous structure and elastic wave velocity measurement of SiO2 glass at high pressures and high temperatures in a Paris-Edinburgh cell

An integration of multi-angle energy-dispersive x-ray diffraction and ultrasonic elastic wave velocity measurements in a Paris-Edinburgh cell enabled us to simultaneously investigate the structures and elastic wave velocities of amorphous materials at high pressure and high temperature conditions. We report the first simultaneous structure and elastic wave velocity measurement for SiO(2) glass at pressures up to 6.8 GPa at around 500°C. The first sharp diffraction peak (FSDP) in the structure factor S(Q) evidently shifted to higher Q with increasing pressure, reflecting the shrinking of intermediate-range order, while the Si-O bond distance was almost unchanged up to 6.8 GPa. In correlation with the shift of FSDP position, compressional wave velocity (Vp) and Poisson's ratio increased markedly with increasing pressure. In contrast, shear wave velocity (Vs) changed only at pressures below 4 GPa, and then remained unchanged at ~4.0-6.8 GPa. These observations indicate a strong correlation between the intermediate range order variations and Vp or Poisson's ratio, but a complicated behavior for Vs. The result demonstrates a new capability of simultaneous measurement of structures and elastic wave velocities at high pressure and high temperature conditions to provide direct link between microscopic structure and macroscopic elastic properties of amorphous materials.