A study on modeling of human spatial behavior using multi-agent technique

This research follows the design and implementation of an agent-based modeling environment written in Java program language on AnyLogic simulation platform to facilitate observing the human spatial behaviors of electric taxis and passengers. The system is developed with a view to improve decision analysis and support for users or companies. In order to achieve the object of decision support for the user, we provide a graphic user interface (GUI) to support user making decision real time. User can simulate the real condition via GUI in our electric taxi DAR operation system to observe the simulation process and the result to help reasonable decision-making immediately. We validate our multi-agent simulation model with the electric taxi DAR operation system case study. Finally, according to the analysis result, we demonstrate that our multi-agent simulation model and GUI can help users or companies quickly make a quality and accurate decision to reduce the decision-making cost and time. In this condition, users or companies can strengthen their competitive advantage and response ability with the reduction of decision-making risk.     

设计与制造之间的界面优化

分辨率增强技术（RET）设计流程中聚集了各式各样的点工具和方法。当前对深亚波长可制造性（DFM）设计的要求迫使设计与制造行业必须建立十分密切的合作关系。同时也促使EDA行业提供一种平台，以促进这些行业间的密切交流。合理利用这一平台将会对产能和品质产生直接的影响。     

Approximate inference in Boltzmann machines

Inference in Boltzmann machines is NP-hard in general. As a result approximations are often necessary. We discuss first order mean field and second order Onsager truncations of the Plefka expansion of the Gibbs free energy. The Bethe free energy is introduced and rewritten as a Gibbs free energy. From there a convergent belief optimization algorithm is derived to minimize the Bethe free energy. An analytic expression for the linear response estimate of the covariances is found which is exact on Boltzmann trees. Finally, a number of theorems is proven concerning the Plefka expansion, relating the first order mean field and the second order Onsager approximation to the Bethe approximation. Experiments compare mean field approximation, Onsager approximation, belief propagation and belief optimization.     

挖掘回归类的混合模型的可识别性

复杂海量数据往往表现为多种结构特征的混合体，回归类混合模型就是对这种混合体的一个描述．该文基于统计学的有限混合分布理论和可识别性的相关结果，针对回归变量的三种情形：(1)解释变量固定，(2)解释变量随机，(3)解释变量固定且类别参数指定，分别讨论挖掘一般回归类的混合模型的可识别性问题，并给出同族回归类混合模型可识别的相应充分条件．这些条件的一个共同特点是它们都与一类特别的解释变量集合有关，而该类集合是由同族的回归函数与回归参数唯一确定的，其元素使不同的回归参数对应回归函数的相同值．特别地，当回归函数线性时，这类集合就是解释变量空间中的超平面．     

A study on information security management system evaluation—assets, threat and vulnerability

The security of information system is like a chain. Its strength is affected by the weakest knot. Since we can achieve 100% Information Security Management System (ISMS) security, we must cautiously fulfill the certification and accreditation of information security. In this paper, we analyzed, studied the evaluation knowledge and skills required for auditing the certification procedures for the three aspects of ISMS—asset, threat, and vulnerability.     

GENERALISED TRANSFER FUNCTIONS OF NEURAL NETWORKS

When artificial neural networks are used to model non-linear dynamical systems, the system structure which can be extremely useful for analysis and design, is buried within the network architecture. In this paper, explicit expressions for the frequency response or generalised transfer functions of both feedforward and recurrent neural networks are derived in terms of the network weights. The derivation of the algorithm is established on the basis of the Taylor series expansion of the activation functions used in a particular neural network. This leads to a representation which is equivalent to the non-linear recursive polynomial model and enables the derivation of the transfer functions to be based on the harmonic expansion method. By mapping the neural network into the frequency domain information about the structure of the underlying non-linear system can be recovered. Numerical examples are included to demonstrate the application of the new algorithm. These examples show that the frequency response functions appear to be highly sensitive to the network topology and training, and that the time domain properties fail to reveal deficiencies in the trained network structure.     

Continuous Cell Separation Using Dielectrophoresis through Asymmetric and Periodic Microelectrode Array

The study presents a dielectrophoretic cell separation method via three-dimensional (3D) nonuniform electric fields generated by employing a periodic array of discrete but locally asymmetric triangular bottom microelectrodes and a continuous top electrode. Traversing through the microelectrodes, heterogeneous cells are electrically polarized to experience different strengths of positive dielectrophoretic forces, in response to the 3D nonuniform electric fields. The cells that experience stronger positive dielectrophoresis are streamed further in the perpendicular direction to the fluid flow, leaving the cells that experience weak positive dielectrophoresis, which continue to traverse the microelectrode array essentially along the laminar flow streamlines. The proposed method has achieved 87.3% pure live cells harvesting efficiency from a live/dead NIH-3T3 cells mixture, and separation of MG-63 cells from erythrocytes with a separation efficiency of 82.8%. The demonstrated cell separation shows promising applications of the DEP separator for cell separation in a continuous mode.     

Linear response algorithms for approximate inference in graphical models

Belief propagation (BP) on cyclic graphs is an efficient algorithm for computing approximate marginal probability distributions over single nodes and neighboring nodes in the graph. However, it does not prescribe a way to compute joint distributions over pairs of distant nodes in the graph. In this article, we propose two new algorithms for approximating these pairwise probabilities, based on the linear response theorem. The first is a propagation algorithm that is shown to converge if BP converges to a stable fixed point. The second algorithm is based on matrix inversion. Applying these ideas to gaussian random fields, we derive a propagation algorithm for computing the inverse of a matrix.     

Electronic band structures and optical gain spectra of strainedwurtzite GaN-AlxGa1-xN quantum-well lasers

The electronic band structures, density-of-states, and optical gain spectra for wurtzite GaN-Al_xGa_1-xN quantum wells are studied theoretically based on the Hamiltonian derived using the k.p method. We investigate the dependence of the optical gain and transparent current density on the well width, barrier height, and strain using a numerical approach with high accuracy. The mole fraction of Al in the barrier material is progressively increased to study the effects of quantum confinement and compressive strain. A higher Al mole fraction in the barrier leads to improvement of the TE optical gain and suppression of the TM optical gain. Furthermore, we demonstrate that a reduction of the well width offers improved modal gain over all radiative current densities. We also predict a transparent current density of 250 A/cm² for the GaN-Al_xGa_1-x N single quantum-well (QW) structure. Our results suggest that a suitable combination of thin well width and large barrier height should be selected in improving the TE optical gain in wurtzite GaN-Al_x Ga_1-xN single QW