A collaborative filtering recommendation algorithm based on item and cloud model

Recommender system is an important content in the research of E-commerce technology. Collaborative filtering recom-mendation algorithm has already been used successfully at recom-mender system. However,with the development of E-commerce,the difficulties of the extreme sparsity of user rating data have become more and more severe. Based on the traditional similarity measuring methods,we introduce the cloud model and combine it with the item-based collaborative filtering recommendation algorithms. The new collaborative filtering recommendation algorithm based on item and cloud model (IC-Based CF) computes the similarity de-gree between items by comparing the statistical characteristic of items. The experimental results show that this method can improve the performance of the present item-based collaborative filtering algorithm with extreme sparsity of data.     

Intelligent outdoor video advertisement recommendation system based on analysis of audiences' characteristics

An integrated implementation framework of an intelligent recommendation system for outdoor video advertising is proposed,which is based on the analysis of audiences' characteristics.Firstly,the images of the scene and the people who view the video advertisements are captured by the network camera deployed on the video advertising terminal side.Then audiences' characteristics can be obtained by applying computer vision technologies:face detection,face tracking,gender recognition and age estimation.Finally,an intelligent recommendation algorithm is designed to decide the most fitting video ads for each terminal according to multi-dimensional statistical information of its audiences' characteristics.The experimental results show that the proposed system can effectively improve the audience arrival rate of the video advertisements by an average growth of 27.04%.Moreover,a novel face detection method and a new face tracking method have been proposed to meet the practical requirements of the system,of which the average F1-score is 0.988 and 0.951 respectively.     

Sodium fluorescence Doppler lidar to measure atmospheric temperature in the mesopause region

A sodium fluorescence Doppler lidar system has been developed to measure environmental parameters of the middle and upper atmosphere. The lidar system mainly comprises a transmitter system,receiver system,data acquisition and control system and data analysis system. A narrowband 589 nm laser is used to excite sodium atoms in the mesopause region. Excitation of the sodium atoms results in resonance fluorescence,which is collected by the receiver. The temperatures in the mesopause region(about 75-105 km) can be derived by analyzing the Doppler-broadened width of the sodium fluorescence. Observations were made with the lidar system,and the number density of sodium atoms and atmospheric temperature profiles were extracted from the observation data. Comparisons of the lidar temperatures and TIMED/SABER temperatures show good agreement,illustrating the reliability of the sodium fluorescence Doppler lidar measurements.     

New method of image retrieval based on interest image region blocks and comprehensive features

We propose an image retrieval method based on interest image region by asymmetrical blocking. An image is segmented into the interest region and background region on a certain rule. For the interest image regions, the color histogram of the uneven blocks is extracted as the color characteristic. We also collect the mean and variance value of the Gabor filtering results of background blocks as texture features of the background image. Then, the images can be retrieved by synthesizing the image color and texture features. We test our approaches by analyzing the results of recall and precision indicators for the Corel image data-base. The experiment results show that the proposed method performs effectively and accurately, which is more effective to retrieve the distant-view images, and the achieved precision increases by about 10% without loss of the retrieval call compared with some other traditional search methods.     

Absolute ruin problems for the risk processes with interest and a constant dividend barrier

In this paper,the absolute ruin in the compound Poisson risk model with interest and a constant dividend barrier is investigated.First,integro-differential equations satisfied by the expected discounted dividend payments are derived.The explicit expressions are obtained when the individual claim size is expo-nential distributed.Second,the moment generating function of the discounted dividends is considered,and integro-differential equations satisfied by the moment generating function of the discounted dividends are derived.Third,by a "differential" argument,the time to recovery to zero from a given negative surplus is considered.Finally,how long it takes for the surplus process to reach the dividend barrier is discussed.     

实验实现光学量子棘轮（英文）

粒子在外势场中的演化可以用薛定谔方程描述.研究发现可以用光场的相位调制模拟粒子的外势场,只要制备出感兴趣的位相分布调制,就可以利用光束的演化模拟粒子在外势场中的演化.实验上利用相位镜产生锯齿状的外势,从而在全光系统中实现了量子棘轮,用肉眼就可以看到22步的棘轮效应.进一步还演示了棘轮效应中的量子共振现象.本工作实现了用经典光模拟量子粒子的演化,从而可以研究其他有趣的量子现象.     

随机利率和随机波动率下保险和再保险公司的稳健最优投资策略 （ 英文 ）

A robust optimal proportional reinsurance and investment management problem is studied fora general insurance company which holds shares of an insurance company and a reinsurance company. Boththe insurance company and reinsurance company are allowed to invest in a risk-free asset and a risky assetwith stochastic interest rate and stochastic volatility, in which the interest rate is described by an affinemodel. Besides, the general insurance company's manager is an ambiguity-averse manager who worriesabout model uncertainty. By employing the dynamic programming approach, the explicit formulae is derivedfor the optimal robust reinsurance-investment strategy and the optimal value function, and one special caseis discussed subsequently. Finally, a part of numerical examples is presented to illustrate the effects ofmodel parameters.     

Experimental demonstrations of high-Q superconducting coplanar waveguide resonators

We successfully designed and fabricated an absorption-type of superconducting coplanar waveguide (CPW) resonators. The resonators are made from a niobium film (about 160 nm thick) on a high-resistance Si substrate, and each resonator is fabricated as a meandered quarter-wavelength transmission line (one end is short to the ground and another end is capacitively coupled to a through feedline). With a vector network analyzer we measured the transmissions of the applied microwave through the resonators at ultra-low temperature. The obtained loaded quality factors are significantly high, i.e. up to ～10 6 . When the temperature increases slowly from the base temperature (20 mK), the resonance frequencies of the resonators are blue shifted and the quality factors are lowered slightly. In principle, this type of device can integrate a series of CPW resonators with a common feedline, making it a promising candidate as the data bus for coupling distant solid-state qubits and the sensitive detector of single photons.     

基于兴趣度的聚类协同过滤推荐系统的设计

协同过滤技术被成功地应用于个性化推荐系统中.随着用户数目和网页数目的日益增加,整个用户矩阵数据极端稀疏并且实时性效果不理想.传统的推荐方法解决不了这些问题.本文结合兴趣度和聚类技术对客户的个人兴趣进行评价,提出了基于兴趣度的聚类协同过滤推荐系统,实验表明,该算法能够有效避免传统方法带来的弊端,提高系统的推荐质量.     

基于Web2．0用户个性化推荐系统分析

为实现用户个性化服务,理解用户兴趣爱好．通过建立用户兴趣模型和推荐库．采用用户兴趣行为描述、重排序算法以及用户反馈算法,分析基于Web2．0用户个性化推荐系统．以提高推荐结果的准确性。     

景区智能空间下的实时景点推荐技术

针对景区景点推荐不能满足游客个性化及实时性的要求,在景区智能空间条件下提出一种实时景点推荐技术,该技术通过上下文感知来获取动态信息,结合景点实时个性化推荐算法为游客提供景点推荐服务.通过实验说明所提出的实时景点推荐技术的可行性.     

一种适合于分布式虚拟环境的兴趣管理方法

为了有效地降低分布式虚拟环境的网络通信量，提出了一种可扩展的兴趣管理方法．该方法采用基于表达式的兴趣表示模型，引入了兴趣度的概念，根据发布一订购模式的过滤机制，实现对信息接收和发送的精度和频率的控制．基于自主开发的分布式虚拟环境原型系统AIMNET的运行结果表明，该方法可以根据对象间的不同兴趣度控制通讯细节，减少了不必要的网络通讯，从而减轻系统的网络负载，进一步提高分布式虚拟环境的可扩展性．     

基于时空序列模式匹配的兴趣点推荐方法

兴趣点(point of interest,POI)是地理空间信息的重要组成部分,在基于位置的信息服务中被广泛使用。针对用户对兴趣点访问所产生的时空序列,利用闭合序列模式挖掘方法分析频繁模式,在此基础上根据用户当前所处位置或最近访问序列,通过序列分析进行时空序列模式匹配,并按照序列的匹配程度给出兴趣点推荐列表。实验结果表明,闭合序列模式挖掘与时空序列模式匹配相结合的方法能够有效地应用在兴趣点推荐中,有利于引导用户的兴趣点访问行为,从而提升位置服务的质量。     

基于用户浏览行为挖掘的电子商务个性化推荐系统

本文介绍了电子商务个性化推荐系统，在此基础上提出了一种新的基于用户浏览行为进行数据挖掘的个性化推荐系统模型，并给出了详细的挖掘流程及实现算法。     

基于页面聚类的个性化推荐算法研究

聚类分析算法作为一种主要的Web使用挖掘技术，在个性化推荐系统中得到了广泛应用，然而面对Web动态性所引起的网页的更新以及用户行为方式的改变，已有的聚类算法并不能很好地解决这一问题。针对这一问题，本文以一种无向图的形式表示用户对网站的访问，提出一种可实时反映网站及用户行为变化情况的增量式页面聚类算法，并在页面聚类的基础上提出相应的推荐决策算法动态生成页面推荐。     

一种结合用户相似度的社会化推荐算法

在对现有的概率矩阵分解算法研究的基础上,针对其中只使用评分信息来做预测存在较大误差的问题,提出了一种结合用户相似度的社会化推荐算法(SRUS).首先,以概率矩阵分解算法(PMF)为基础,结合用户相似度信息进行建模;其次,使用潜在特征空间将评分矩阵和相似度矩阵关联到一个统一的框架中;最后,对这2个矩阵进行矩阵分解,实现算法的优化推荐.将这一算法与PMF算法进行比较,实验表明,SRUS算法在数据稀疏性、冷启动和精确性方面具有更优的效果.     

一种基于协同推荐的网页排序算法

针对目前搜索结果个性化排序算法中的用户兴趣模型构建难、相关度计算不精确等问题,提出了一种结合用户兴趣模型和协同推荐算法的个性化排序方法.该方法从用户的搜索历史,包括提交查询、点击相关网页等反馈信息来训练用户的兴趣模型,然后采用协同推荐算法获取具有共同兴趣的邻居用户,根据这些邻居对网页的推荐程度和网页与用户的相关程度来排序搜索结果.实验结果表明:该排序算法的平均最小精确度比一般排序算法提高了约0.1,且随着用户邻居数目的增长,最小精确度随之增长.与其他排序算法相比,采用协同推荐算法有助于提高网页与用户兴趣关联程度计算的精确度,从而提高排序的效率,有助于改善用户的搜索体验.     

基于Web使用挖掘的个性化推荐服务研究

本文研究结合网站结构和页面内容以辅助Web使用挖掘,提高推荐服务质量。并在相关理论与研究的基础上,给出一种关于推荐因子的综合计算方法,经推荐质量分析,该方法具有较好的推荐优化能力。