Gaussian process volatility model

The prediction of time-changing variances is an important task in the modeling of financial data. Standard econometric models are often limited as they assume rigid functional relationships for the evolution of the variance. Moreover, functional parameters are usually learned by maximum likelihood, which can lead to over-fitting. To address these problems we introduce GP-Vol, a novel non-parametric model for time-changing variances based on Gaussian Processes. This new model can capture highly flexible functional relationships for the variances. Furthermore, we introduce a new online algorithm for fast inference in GP-Vol. This method is much faster than current offline inference procedures and it avoids overfitting problems by following a fully Bayesian approach. Experiments with financial data show that GP-Vol performs significantly better than current standard alternatives.

Behavior model of a CMOS process compatible photo-diode

A behavior model of a photo-diode is presented. The model describes the relationship between photocurrent and incident optical power, and it also illustrates the impact of the reverse bias to the variation of the junction capacitance. With this model, the photo-diode and a CMOS receiver circuit were simulated and designed simultaneously under a universal circuit simulation environment.

a flexible negotiation model for an agent-based software process modelling

Our Agent-based Software Process Modelling (ASPM) approach describes a software process as a set of cooperative agents. Negotiation is the way in which the agents construct their cooperative relations, and thus the software process. Currently, most negotiation models use a fixed negotiation protocol and fixed strategies. In order to achieve the flexibility that the negotiation of the agents in ASPM requires, we propose a negotiation model NM-PA. NM-PA mainly includes a generic negotiation protocol and some rules, which possibly change in different negotiation processes. By changing the rules, the model can support multi-protocols and multi-decision-making strategies at a lower cost.

the trustworthiness metric model of software process quality based-on life circle

Communication University of China; Wuhan University; James Madison University; Institute of Policy and Management, Chinese Academy of Sciences; IEEE Wuhan Section

Strangeness production process pp -> nK(+)Sigma(+) within resonance model

We explore production mechanism and final state interaction in the pp -> nK(+)Sigma(+) channel based on the inconsistent experimental data published respectively by COSY-11 and COSY-ANKE. The scattering parameter a > 0 for n Sigma(+) interaction is favoured by large near-threshold cross section within a nonrelativistic parametrization investigation, and a strong n Sigma(+) interaction comparable to pp interaction is also indicated. Based on this analysis we calculate the contribution from resonance Delta*(1920) through pi(+) exchange within resonance model, and the numerical result suggests a rather small near-threshold total cross section, which is consistent with the COSY-ANKE data. With an additional sub-threshold resonance Delta*(1620), the model gives a much better description to the rather large near-threshold total cross section published by COSY-11

多分量对偶超导理论和口袋模型；Multi-component Dual Superconductor Theory and Bag Model

基于联络新参数化方案研究了多分量对偶超导模型。给出了多分量Ginzburg-Landau模型中的自对偶解,并研究了磁通量子数趋于无穷大时的墙涡旋解,以及与口袋模型之间的联系。

the study of trustworthy software process improvement model

Huazhong Univ Sci & Technol, Natl Tech Univ Ukraine, Huazhong Normal Univ, Harbin Inst Technol, IEEE Ukraine Sect, I& M/CI Joint Chapter

Supervised Gaussian Process Latent Variable Model for Dimensionality Reduction

The Gaussian process latent variable model (GP-LVM) has been identified to be an effective probabilistic approach for dimensionality reduction because it can obtain a low-dimensional manifold of a data set in an unsupervised fashion. Consequently, the GP-LVM is insufficient for supervised learning tasks (e. g., classification and regression) because it ignores the class label information for dimensionality reduction. In this paper, a supervised GP-LVM is developed for supervised learning tasks, and the maximum a posteriori algorithm is introduced to estimate positions of all samples in the latent variable space. We present experimental evidences suggesting that the supervised GP-LVM is able to use the class label information effectively, and thus, it outperforms the GP-LVM and the discriminative extension of the GP-LVM consistently. The comparison with some supervised classification methods, such as Gaussian process classification and support vector machines, is also given to illustrate the advantage of the proposed method.

On the Behavior of the Homogeneous Self-Dual Model for Conic Convex Optimization

There is a natural norm associated with a starting point of the homogeneous self-dual (HSD) embedding model for conic convex optimization. In this norm two measures of the HSD model’s behavior are precisely controlled independent of the problem instance: (i) the sizes of ε-optimal solutions, and (ii) the maximum distance of ε-optimal solutions to the boundary of the cone of the HSD variables. This norm is also useful in developing a stopping-rule theory for HSD-based interior-point methods such as SeDuMi. Under mild assumptions, we show that a standard stopping rule implicitly involves the sum of the sizes of the ε-optimal primal and dual solutions, as well as the size of the initial primal and dual infeasibility residuals. This theory suggests possible criteria for developing starting points for the homogeneous self-dual model that might improve the resulting solution time in practice

Towards a new model of criminal justice system in the era of globalised criminality: the biggest challenges for criminal process legislation

The main objective of this thesis is the critical analysis of the evolution of the criminal justice systems throughout the past decade, with special attention to the fight against transnational terrorism. It is evident – for any observer - that such threats and the associated risk that terrorism entails, has changed significantly throughout the past decade. This perception has generated answers – many times radical ones – by States, as they have committed themselves to warrant the safety of their populations and to ease a growing sentiment of social panic. This thesis seeks to analyse the characteristics of this new threat and the responses that States have developed in the fight against terrorism since 9/11, which have questioned some of the essential principles and values in place in their own legal systems. In such sense, freedom and security are placed into perspective throughout the analysis of the specific antiterrorist legal reforms of five different States: Israel, Portugal, Spain, the United Kingdom and the United States of America. On the other hand, in light of those antiterrorist reforms, it will be questioned if it is possible to speak of the emergence of a new system of criminal justice (and of a process of a convergence between common law and civil law systems), built upon a control and preventive security framework, significantly different from traditional models. Finally, this research project has the fundamental objective to contribute to a better understanding on the economic, social and civilization costs of those legal reforms regarding human rights, the rule of law and democracy in modern States.