随机森林理论浅析

随机森林是一种著名的集成学习方法,被广泛应用于数据分类和非参数回归。本文对随机森林算法的主要理论进行阐述,包括随机森林收敛定理、泛化误差界以和袋外估计三个部分。最后介绍一种属性加权子空间抽样的随机森林改进算法,用于解决超高维数据的分类问题。     

Development of the heat treatment system for the 40 T hybrid magnet superconducting outsert

The heat treatment of Nb(3)Sn coil with the glass fabric insulation is one of the key and critical processes for the outsert solenoids of the 40 T hybrid magnet, which could be wound with cable-in-conduit conductors using the insulation-wind-and-react technique. The manufacturing of the large vertical type vacuum/Ar atmosphere-protection heat treatment system has been completed and recently installed in the High Magnetic Filed Laboratory, Chinese Academy of Sciences. The heat treatment system composed mainly the furnace, the purging gas supply system, the control system, the gas impurities monitoring system, and so on. At present, the regulation and testing of the heat treatment system has been successfully finished, and all of technical parameters meet or exceed specifications.     

Self-eliminating instrumental frequency response from free carrier absorption signals for silicon wafer characterization

Accurate determination of electronic transport properties of semiconductor wafers with modulated free carrier absorption (MFCA) and multiparameter fitting requires the total elimination of instrumental response from the MFCA signals. In this paper, an approach to eliminate the effect of instrumental response on the frequency dependence of MFCA amplitude and phase is developed both theoretically and experimentally to simultaneously determine the transport properties (minority-carrier lifetime, carrier diffusion coefficient, and front surface recombination velocity) of silicon wafers. Experimental results showed that with the proposed method the instrumental frequency response was fully eliminated from the experimental MFCA data and had no impact on the multiparameter fitting, while with conventional methods the accuracy of the fitted transport parameters was influenced detrimentally by the errors of the measured instrumental frequency responses, in particular for the minority-carrier lifetime and the front surface recombination velocity.     

基于ajax+php技术的图书馆电子光盘管理系统

该套系统使用Ajax+Php技术,使得图书管理员通过可视化管理界面,将大容量的图书光盘资料上传至后台磁盘柜中,并将图书的数据信息输入到数据库中,为师生提供了更加便捷的查阅资料方式。重点对系统使用技术进行详细介绍。     

中职《计算机网络》课程实验教学的建设与实践

随着科技的进步和网络的日益普及,《计算机网络》成为中职计算机专业的核心课程,所以,《计算机网络》课程的教学需要不断完善教学体系,重视建设有效的实验教学形式与方法,才能培养出高素质实践型的计算机人才。     

Technology-Based Service Encounters Using Self-Service Technologies in the Healthcare Industry

Although there have been studies discussing the influence of technology-based services on the overall service efficiency and quality of organizations in various industries, very little effort has been devoted to investigating this issue in the healthcare industry. Hospital image is considered to be a crucial factor influencing patients' choice of hospitals, but few studies specifically examine its association with technology-based services. By consulting the model of the European Customer Satisfaction Index, a research model for evaluating the impact of the use of technology-based services on hospital image, patient satisfaction, and patient loyalty in the healthcare industry is developed and examined in this study using survey data collected from 738 patients at two medical centers with an online appointment system. The research results confirm the importance of providing quality, technology-based services in enhancing hospital image, patient satisfaction, and patient loyalty. The implications of this research and suggestions for future work are also discussed.     

Adaptive detection of range-spread targets without secondary data in multichannel autoregressive process

Adaptive detection of range-spread targets without secondary data is addressed in a multichannel autoregressive Gaussian disturbance with unknown space–time covariance matrix, by utilizing the Rao test. The proposed Rao test without secondary data is theoretically proved to be asymptotically (large-sample in the number of temporal observations) constant false alarm rate with respect to unknown space–time covariance matrix, thanks to an asymptotic equivalence between the Rao test and the generalized likelihood ratio test. Moreover, the performance loss due to no secondary data can be remedied by appropriately increasing the temporal dimension. The performance assessment conducted by Monte Carlo simulation, also in comparison with the existing detector without secondary data, confirms the effectiveness of the proposed detectors.     

Bare finger 3D air‐touch system with embedded multiwavelength optical sensor arrays for mobile 3D displays

A camera‐free 3D air‐touch system was proposed. Hovering, air swiping, and 3D gestures for further interaction with the floated 3D images on the mobile display were demonstrated. By embedding multiwavelength optical sensors into the display pixels and adding angular‐scanning illuminators with multiwavelength on the edge of the display, the flat panel can sense images reflected by a bare finger from different heights. In addition, three axis (x, y, z) information of the reflected image of the fingertip can be calculated. Finally, the proposed 3D air‐touch system was successfully demonstrated on a 4‐inch mobile 3D display.     

A study of optimal viewing distance in a parallax barrier 3D display

The optimal viewing distance was proposed as a parameter for designing a parallax barrier 3D display. It can be designed based on simple geometric method and by considering the pitches of image display pixels and parallax barrier, or even including the aperture ratios of the image display pixels and parallax barrier. It can be analyzed by using ray tracing method. By considering the optical refraction index of the cover glass, the angular behavior of the system becomes more realistic; however, the geometric method is difficult to be used. We propose a revised method for estimating the view distance and angular behavior. In this paper, we have demonstrated a designated eye position (DEP) for each view and shown that multiple DEPs make a circular curve around the center of the display. We prove the new concept by comparing the optical ray tracing calculations and optical measurement.     

Extension of the gap statistics index to fuzzy clustering

The well-known gap statistic index proposed by Tibshirani et al. has successfully applied in many clustering evaluations. However, the gap statistic index cannot evaluate the clustering partitions from any fuzzy clustering algorithm. This is because fuzzy clustering cannot provide the within-cluster similarity measure that is used in the gas statistic index. Thus, the applicable range of the gap statistic index is very limited. In this paper, we present a new method that extends the gap statistic index to fuzzy clustering by using fuzzy membership notations. Our proposed method can extend the applicability of the gap statistic index, and outperform other existing fuzzy indices in several aspects. Experiments on eight sets of synthetic and real datasets are used to verify the applicability and efficiency of the proposed method.