大数据量图书下多数据集的二部图多样化推荐

文章利用加权二部图推荐算法在电子商务产品多样化推荐中的优势,并根据图书的特点对以往推荐方法进行相应改进。提出利用spider从当当网抓取大数据量的图书购买记录,结合高校图书馆借阅记录,利用用户对图书的评分作为二部图的权值。综合用户之间对不同图书评分的偏好预测、对相同图书评分的偏好预测,以及借阅偏好预测,进行图书推荐。在两种数据集上,通过实验获取最佳的实验参数,验证提出的方法不但能推荐多样化的图书,推荐准确率也得到提高。

多数据集深度学习模型的修图处理识别

图像处理软件的飞速发展,带动了移动应用领域一大批修图、美化应用的兴起。但是修图、美化软件的快速发展和普及也带来了一些社会问题和安全问题,如网恋对象严重失真,摄影作品造假等。针对手机中的修图处理APP软件,提出一种基于多数据集特征学习的神经网络模型,并给出其网络拓扑结构。区别于传统的多个神经网络并行操作,提出的网络模型具有共享模型参数的特征,能同时对多个特征数据集进行深度学习,使检测程序具备多特征识别能力。此外,还提出了一种针对多任务网络模型的损失函数,以增强深度特征学习的能力。实验结果表明,提出方法的准确率较传统方法有较大提升,同时泛化性能优越,能识别出经过多种美图、修图软件修复过的图像。

基于多数据集动态潜变量的在线性能分级评估方法

针对动态多变量过程中难以提取明确的过程变量的动态关系问题,本文提出基于多数据集动态潜变量分析(MSDLV)的在线性能分级评估的方法.首先将性能相近的过程历史数据段划分为不同性能等级的集合,然后运用MSDLV方法提取性能级之间的公共基向量,保留训练数据中性能相关的过程变化,将性能相关的特有变化分解为动态部分与静态部分,提取动态自相关过程的动态因素.建立动态潜变量与性能等级之间的离线模型后,在线评估当前过程性能以及判断其所处状态.最后,将该方法运用于乙烯裂解炉反应过程,结果表明该方法具有良好的准确度.

基于SVM与多数据集的摔倒检测方法研究

摔倒检测系统能够实时监测用户的运动状态,当用户摔倒时,自动发出警报,让用户得到及时的救治。对于摔倒检测方法的研究通常采用自建数据集进行验证实验,然而,数据集的差异影响实验结果的好坏,实验结果无法进行比较,影响实验结果的可信度。文章使用三个基于加速度数据的摔倒检测公开数据集,采用基于运动方向特征及支持向量机(SVM)的方法进行分类实验。分类算法在三个数据集上的实验结果为:Mobi Act数据集的灵敏度与特异性为97.41%和98.10%;t Fall数据集的灵敏度与特异性为96.55%和97.86%;Sis Fall数据集的灵敏度与特异性为98.77%和99.03%。

基于加权的多数据集时态关联规则算法

文章将子数据集中的时态关联规则的时态因素进行了区间归一和延伸,再参考有效投票率进行规则剪辑,对剪辑后的子数据库中的规则进行加权计算得到最终的时态关联规则,最后通过算法伪代码具体实现了该算法。

异质数据集关联规则挖掘

针对存在的关联规则挖掘算法不能有效地在异质数据集中进行,本文首先使用领域本体方法处理数据集中的异质现象,然后提出了一种有效的XML异质数据集关联规则挖掘算法,实验结果表明该算法在挖掘速度和挖掘时在对内存的占用方面都优于现有的算法。

基于MapReduce的多路连接优化方法研究

MapReduce是Google开发的一种并行分布式计算模型,已在搜索和处理海量数据领域得到了广泛的应用。但是MapReduce框架中的"一对一分区"策略使得其在处理多数据连接任务时,需要将该任务拆分成多个链接的子任务,造成中间结果集的频繁"洗牌",带来巨大的磁盘I/O开销。文中就该问题提出了一种新的分区策略:"一对多分区"策略,为了能够在MapReduce框架中实现这一分区策略,因此需要对MapReduce框架中的分区函数接口进行修改。改进策略的优点在于只要一个MapReduce任务就能够完成多数据集连接任务,因此节省了I/O开销。最后在搭建的Hadoop平台上对改进前和改进后的两种方法进行比较。实验结果表明,改进模式的效率明显得到提高,因此这一方案是可行的。

广义多线性混合效应模型

为了解决多数据集间联合特征提取时数据分布多样、集间相关性结构复杂和共享特征方法多样的问题,本文提出了广义多线性混合效应模型。作为一种非监督多数据集特征提取方法,本算法可挖掘多个数据集之间的共享信息,实现对多数据集全局、局部和个体特征的提取。本算法利用了传统的广义线性模型,使其可以处理不同分布的张量数据集;并提出了一种基于超图的关系模型。该模型利用关系矩阵可以实现对数据集间相关结构的建模;通过提出辅助模式的概念,实现了特征的自动归类。数值实验结果表明:利用本算法提取的特征不仅反映了多数据集间的共同与个体信息,并且在人脸识别和推荐系统等问题中性能优于传统算法。

Design of multi-channel data acquisition system based on LabVIEW and NI PXI-5105

A data acquisition system based on LabVIEW and NI PXI-5105 is presented for multi-channel data acquisition. It can realize the functions of parameter setting, data acquisition and storage, waveform display and data analysis using LabVIEW and NI-SCOPE device driver. The advantages of the system are that the setting is convenient, the operation is easy, the interface is friendly and the functions are practical. The experiment results show that the system has good stability and high reliability and is a powerful tool for multi-channel data acquisition.

H.264编码器的SSE2指令级优化

H.264视频编码标准采用了很多新技术,具有更优越的编码效率,同时也增加了计算复杂度,无法满足实时应用。由于单指令多数据扩展指令集2(SSE2)的并行运算能力可以提高计算机对多媒体数据的实时处理。文中主要采用了SSE2对H.264中的一些耗时较多的关键模块,例如整数像素运动估计中计算SAD、整数DCT变换、量化、Hadamard变换以及亚像素运动估计中计算SATD进行了指令级优化。实验结果表明,经过优化后,在保持视频图像质量的前提下,相应模块运行速度得到了提高,使H.264编码器整体的编码速度较好地满足实时要求。

等效环路时域有限差分算法及其核内加速技术

提出了一种新颖的有限差分算法——等效环路有限差分算法。这种算法借鉴传输线算法的思想,在Yee氏网格中引入集总元件的概念,由于其良好的物理思想,可实现对于色散介质的仿真计算。单指令多数据指令集(streaming single-instruction multiple-data extensions,SSE)技术是Intel公司提出的核内并行加速技术,普通的个人电脑都支持这种技术。结合SSE技术,可实现对算法的有效加速,这种加速技术无须硬件的改进即可实现理论上最高4倍的加速。分别利用等效环路有限差分算法、普通有限差分算法仿真计算了平面倒F天线(pla-nar inverted-F antenna,PIFA)的端口反射系数和天线方向图,并与实测数据进行了比较,验证了算法的正确性。在实际的运算中,使用SSE技术普遍可实现2倍以上的加速比,加速算法的有效性得到了验证。

Effect of Recycled Coarse Aggregate on Concrete Compressive Strength

The effect of recycled coarse aggregate on concrete compressive strength was investigated based on the concrete skeleton theory. For this purpose, 30 mix proportions of concrete with target cube compressive strength ranging from 20 to 60 MPa were cast with normal coarse aggregate and recycled coarse aggregate from different strength parent concretes. Results of 28-d test show that the strength of different types of recycled aggregate affects the concrete strength obviously. The coarse aggregate added to mortar matrix plays a skeleton role and improves its compressive strength. The skeleton effect of coarse aggregate increases with the increasing strength of coarse aggregate, and normal coarse aggregate plays the highest, whereas the lowest concrete strength occurs when using the weak recycled coarse aggregate. There is a linear relationship between the concrete strength and the corresponding mortar matrix strength. Coarse aggregate skeleton formula is established, and values from experimental tests match the derived expressions.

Similarity measure design for high dimensional data

Information analysis of high dimensional data was carried out through similarity measure application. High dimensional data were considered as the a typical structure. Additionally, overlapped and non-overlapped data were introduced, and similarity measure analysis was also illustrated and compared with conventional similarity measure. As a result, overlapped data comparison was possible to present similarity with conventional similarity measure. Non-overlapped data similarity analysis provided the clue to solve the similarity of high dimensional data. Considering high dimensional data analysis was designed with consideration of neighborhoods information. Conservative and strict solutions were proposed. Proposed similarity measure was applied to express financial fraud among multi dimensional datasets. In illustrative example, financial fraud similarity with respect to age, gender, qualification and job was presented. And with the proposed similarity measure, high dimensional personal data were calculated to evaluate how similar to the financial fraud. Calculation results show that the actual fraud has rather high similarity measure compared to the average, from minimal 0.0609 to maximal 0.1667.

Effect of multidisciplinary team treatment on outcomes of patients with gastrointestinal malignancy

AIM: To evaluate the effect of multidisciplinary team (MDT) treatment modality on outcomes of patients with gastrointestinal malignancy in China. METHODS: Data about patients with gastric and colorectal cancer treated in our center during the past 10 years were collected and divided into two parts. Part 1 consisted of the data collected from 516 consecutive complicated cases discussed at MDT meetings in Peking University School of Oncology (PKUSO) from December 2005 to July 2009. Part 2 consisted of the data collected from 263 consecutive cases of resect-able locally advanced rectal cancer from January 2001 to January 2005. These 263 patients were divided into neoadjuvant therapy (NT) group and control group. Patients in NT group received MDT treatment, namely neoadjuvant therapy + surgery + postoperative adjuvant therapy. Patients in control group underwent direct surgery + postoperative adjuvant therapy. The outcomes in two groups were compared. RESULTS: The treatment strategy was altered after discussed at MDT meeting in 76.81% of gastric cancer patients and in 58.33% of colorectal cancer patients before operation. The sphincter-preservation and local control of tumor were better in NT group than in control group. The 5-year overall survival rate was also higher in NT group than in control group (77.23% vs 69.75%, P = 0.049). CONCLUSION: MDT treatment modality can significantly improve the outcomes of patients with gastrointestinal malignancy in China.

Multiscale Multisensor Data Fusion and Application in High Precision Marking and Cutting Robot System

The multisensor online measure system for high precision marking and cutting robot system is designed and the data fusion method is introduced, which combines augment state multiscale process with extend Kalman filter. The technology measuring the three-dimensional deforming information of profiled bars is applied. The experimental result shows that applying the multisensor data fusion technology can enhance the measure precision and the reliability of measure system.

Weekend and nighttime effect on the prognosis of peptic ulcer bleeding

AIM:To evaluate whether weekend or nighttime admission affects prognosis of peptic ulcer bleeding despite early endoscopy.METHODS:Retrospective data collection from four referral centers,all of which had a formal out-of-hours emergency endoscopy service,even at weekends.A total of 388 patients with bleeding peptic ulcers who were admitted via the emergency room between January 2007 and December 2009 were enrolled.Analyzed parameters included time from patients' arrival until endoscopy,mortality,rebleeding,need for surgery and length of hospital stay.RESULTS:The weekday and weekend admission groups comprised 326 and 62 patients,respectively.There were no significant differences in baseline characteristics between the two groups,except for younger age in the weekend group.Most patients (97%) had undergone early endoscopy,which resulted in a low mortality rate regardless of point of presentation (1.8% overall vs 1.6% on the weekend).The only outcome that was worse in the weekend group was a higher rate of rebleeding (12% vs 21%,P = 0.030).However,multivariate analysis revealed nighttime admission and a high Rockall score (≥ 6) as significant independent risk factors for rebleeding,rather than weekend admission.CONCLUSION:Early endoscopy for peptic ulcer bleeding can prevent the weekend effect,and nighttime admission was identified as a novel risk factor for rebleeding,namely the nighttime effect.

Data acquisition,analysis and applications of multi-sensor integration

This paper presents some key techniques for multi-sensor integration system, which is applied to the intelligent transportation system industry and surveying and mapping industry, e.g. road surface condition detection, digital map making. The techniques are synchronization control of multi-sensor, space-time benchmark for sensor data, and multi-sensor data fusion and mining. Firstly, synchronization control of multi-sensor is achieved through a synchronization control system which is composed of a time synchronization controller and some synchronization sub-controllers. The time synchronization controller can receive GPS time information from GPS satellites, relative distance information from distance measuring instrument and send space-time information to the synchronization sub-controller. The latter can work at three types of synchronization mode, i.e. active synchronization, passive synchronization and time service synchronization. Secondly, space-time benchmark can be established based on GPS time and global reference coordinate system, and can be obtained through position and azimuth determining system and synchronization control system. Thirdly, there are many types of data fusion and mining, e.g. GPS/Gyro/DMI data fusion, data fusion between stereophotogrammetry and PADS, data fusion between laser scanner and PADS, and data fusion between CCD camera and laser scanner. Finally, all these solutions presented in paper have been applied to two areas, i.e. land-bone intelligent road detection and measurement system and 3D measurement system based on unmanned helicopter. The former has equipped some highway engineering Co. , Ltd. and has been successfully put into use. The latter is an ongoing resealch.

Poorly expandable common bile duct with stones on endoscopic retrograde cholangiography

AIM： To describe characteristics of a poorly expandable （PE） common bile duct （CBD） with stones on en- doscopic retrograde cholangiography.METHODS： APE bile duct was characterized by a rigid and relatively narrowed distal CBD with retro- grade dilatation of the non-PE segment. Between 2003 and 2006, endoscopic retrograde cholangiography （ERC） images and chart reviews of 1213 patients with newly diagnosed CBD stones were obtained from the computer database of Therapeutic Endoscopic Centerin Chang Gung Memoria acteristic PE bile duct on Hospital. Patients with char ERC were identified from the database. Data of the patients as well as the safety and technical success of therapeutic ERC were collected and analyzed retrospectively.RESULTS： A total of 30 patients with CBD stones and characteristic PE segments were enrolled in this study. The median patient age was 45 years （range, 20 to 92 years）; 66.7% of the patients were men. The di ameters of the widest non-PE CBD segment, the PE segment, and the largest stone were 14.3 ± 4.9 mm, 5.8±1.6 mm, and 11.2±4.7 mm, respectively. The length of the PE segment was 39.7±15.4 mm （range, 12.3 mm to 70.9 mm）. To remove the CBD stone（s） completely, mechanical lithotripsy was required in 25 （83.3%） patients even though the stone size was not as large as were the difficult stones that have been described in the literature. The stone size and stone/ PE segment diameter ratio were associated with the need for lithotripsy. Post-ERC complications occurred in 4 cases： pancreatitis in 1, cholangitis in 2, and an im- pacted Dormia basket with cholangitis in 1. Two （6.7%） of the 28 patients developed recurrent CBD stones at follow-up （50±14 mo） and were successfully managed with therapeutic ERC.CONCLUSION： Patients with a PE duct frequently require mechanical lithotripsy for stones extraction, To retrieve stones successfully and avoid complications, these patients should be identified during ERC,

Junk band recovery for hyperspectral image based on curvelet transform

Under consideration that the profiles of bands at close wavelengths are quite similar and the curvelets are good at capturing profiles, a junk band recovery algorithm for hyperspectral data based on curvelet transform is proposed. Both the noisy bands and the noise-free bands are transformed via curvelet band by band. The high frequency coefficients in junk bands are replaced with linear interpolation of the high frequency coefficients in noise-flee bands, and the low frequency coefficients remain the same to keep the main spectral characteristics from being distorted. Jutak bands then are recovered after the inverse curvelet transform. The performance of this method is tested on the hyperspectral data cube obtained by airborne visible/infrared imaging spectrometer （AVIRIS）. The experimental results show that the proposed method is superior to the traditional denoising method BayesShrink and the art-of-state Curvelet Shrinkage in both roots of mean square error （RMSE） and peak-signal-to-noise ratio （PSNR） of recovered bands.

Multisource least-squares reverse-time migration with structure-oriented filtering

The technology of simultaneous-source acquisition of seismic data excited by several sources can significantly improve the data collection efficiency. However, direct imaging of simultaneous-source data or blended data may introduce crosstalk noise and affect the imaging quality. To address this problem, we introduce a structure-oriented filtering operator as preconditioner into the multisource least-squares reverse-time migration （LSRTM）. The structure-oriented filtering operator is a nonstationary filter along structural trends that suppresses crosstalk noise while maintaining structural information. The proposed method uses the conjugate-gradient method to minimize the mismatch between predicted and observed data, while effectively attenuating the interference noise caused by exciting several sources simultaneously. Numerical experiments using synthetic data suggest that the proposed method can suppress the crosstalk noise and produce highly accurate images.