Is my model any good: differentially private regression diagnostics

Linear and logistic regression are popular statistical techniques for analyzing multi-variate data. Typically, analysts do not simply posit a particular form of the regression model, estimate its parameters, and use the results for inference or prediction. Instead, they first use a variety of diagnostic techniques to assess how well the model fits the relationships in the data and how well it can be expected to predict outcomes for out-of-sample records, revising the model as necessary to improve fit and predictive power. In this article, we develop \(\epsilon \)-differentially private diagnostics tools for regression, beginning to fill a gap in privacy-preserving data analysis. Specifically, we create differentially private versions of residual plots for linear regression and of receiver operating characteristic (ROC) curves as well as binned residual plot for logistic regression. The residual plot and binned residual plot help determine whether or not the data satisfy the assumptions underlying the regression model, and the ROC curve is used to assess the predictive power of the logistic regression model. These diagnostics improve the usefulness of algorithms for computing differentially private regression output, which alone does not allow analysts to assess the quality of the posited model. Our empirical studies show that these algorithms can be effective tools for allowing users to evaluate the quality of their models.     

Early bankruptcy prediction using ENPC

Bankruptcy prediction has long time been an active research field in finance. One of the main approaches to this issue is dealing with it as a classification problem. Among the range of instruments available, we focus our attention on the Evolutionary Nearest Neighbor Classifier (ENPC). In this work we assess the performance of the ENPC comparing it to six alternatives. The results suggest that this algorithm might be considered a good choice.     

A process mining-based analysis of business process work-arounds

Business process work-arounds are specific forms of incompliant behavior, where employees intentionally decide to deviate from the required procedures although they are aware of them. Detecting and understanding the work-arounds performed can guide organizations in redesigning and improving their processes and support systems. Existing process mining techniques for compliance checking and diagnosis of incompliant behavior rely on the available information in event logs and emphasize technological capabilities for analyzing this information. They do not distinguish intentional incompliance and do not address the sources of this behavior. In contrast, the paper builds on a list of generic types of work-arounds found in practice and explores whether and how they can be detected by process mining techniques. Results obtained for four work-around types in five real-life processes are reported. The remaining two types are not reflected in events logs and cannot be currently detected by process mining. The detected work-around data are further analyzed for identifying correlations between the frequency of specific work-around types and properties of the processes and of specific activities. The analysis results promote the understanding of work-around situations and sources.     

Modelling salient visual dynamics in videos

Automatic video annotation is a critical step for content-based video retrieval and browsing. Detecting the focus of interest in video frames automatically can benefit the tedious manual labeling process. However, producing an appropriate extent of visually salient regions in video sequences is a challenging task. Therefore, in this work, we propose a novel approach for modeling dynamic visual attention based on spatiotemporal analysis. Our model first detects salient points in three-dimensional video volumes, and then uses the points as seeds to search the extent of salient regions in a novel motion attention map. To determine the extent of attended regions, we use the maximum entropy in the spatial domain to analyze the dynamics derived by spatiotemporal analysis. Our experiment results show that the proposed dynamic visual attention model achieves high precision value of 70% and reveals its robustness in successive video volumes.     

A ctivation measurement of neutron distribution in intermediate energy heavy ion target area

The distribution of neutrons produced by the reaction of 50MeV/u ^12C-ion on a thick Cu target are studied.The neutrons are measured with threshold activation detectors.Al.F,C,Al and In activation samples were used to measure neutrons with energy greater than 7,11,20,50MeV and thermal neutrons,respectively,The fluence rate,energy and angular distributions of neutrons,total neutron yield of 12C-ion and the emission rate in the forward direction of neutrons over 11 and 20MeV were obtained.     

Beyond the financial crisis: a way forward with development paradigm

As the impact of the financial crisis spreads worldwide, it has become a top priority of various countries, international institutions, entrepreneurs and scholars to find innovative and creative ways to face this challenge. As Hazel Henderson (2002) has pointed out, “the world has not fallen into a financial crisis, but fundamentally fell into a crisis of development paradigm.” We need to reflect seriously on this paradigm and rethink of the social and economic models and cultural values for meeting the challenges of this crisis. The paper introduces key issues of development within a global context, and explores the shift of the value system from hard technology centered to soft technology-based future direction of human development.     

Linear-Dispersion Division Multiple-Access for MIMO systems

Comprehensive study on novel Linear-Dispersion Division Multiple-Access（LDDMA） for multi-user uplink Multiple-Input Multiple-Output（MIMO）systems is proposed.In the new multiplexing scheme,each user’s information symbol is dispersed by a User-Specific Matrix（USM）both in space and time domain and linearly combined at base-station side.And a simple random search algorithm,based on capacity maximization criteria,is developed to generate a bank of USMs.Simulation results are presented to demonstrate the advantages of LDDMA.When the Bit Error Rate（BER） reaches 10^–3,the performance gains are 3dB and 5dB,compared with Time-Division Linear Dispersion Codes（TD-LDC）and BLAST,respectively.     

基于模糊神经网络的母线充电保护研究

针对当前母线充电试验时继电保护中存在的一些问题,提出了一种智能型的母线充电保护新方法。在分析母线正常充电和故障元件充电的基础上,找到了正确快速地识别两种情况的特征量,结合模糊神经网络这一新型的人工智能技术,综合利用母线充电时的多种电气量分别提取形成网络的特征输入量,并采用Simpson模糊极小-极大神经网络来形成区分母线正常充电和故障元件充电的模糊模式识别器。利用EMTP程序仿真来获取不同系统参数和各种不同故障情况下模糊神经网络训练和测试所需要的大量样本,通过对模糊神经网络的训练,使网络具备了很强的故障识别能力和较强的泛化能力,结果表明,训练后的网络能快速准确地区分母线在各种运行工况下的正常充电和故障元件充电,从而验证了该方法的有效性。     

A New Construction of Signature Waveforms for Multi-rate Multi-cell QS-CDMA Systems

In this paper, we propose a new construction of signature waveform sets based on Generalized Loosely Synchronization (GLS) sets and different chip waveforms. The new signature sets are applied into the multi-rate multi-cell quasi-synchronous CDMA (QS-CDMA) system where each cell is assigned with a GLS set; different users in the same cell are assigned with different GLS sequences in the same GLS set; user’s different streams are assigned with the same GLS sequence but different chip waveforms. According to the properties of GLS sets, the inter-cell interference (ICI) and the multi-user interference (MUI) in the same cell can be reduced significantly. The interferences among different streams of the same user are handled by an optimal (or suboptimal) multi-stream detector(s), notice that the multi-stream detector mentioned here is also named as multi-user detector in other references. We compare the performance of the multi-rate multi-cell QS-CDMA system employing the proposed sets with that of multi-rate system employing well-known concatenated orthogonal/PN sets and that of single-rate system employing GLS sets. The results show that the multi-rate system employing the proposed sets can achieve significant interference reduction. Meanwhile the performance of multi-rate system is similar to that of single-rate system due to the inclusion of multi-user detection.