Dependent mixtures of Dirichlet processes

An approach to modeling dependent nonparametric random density functions is presented. This is based on the well known mixture of Dirichlet process model. The idea is to use a technique for constructing dependent random variables, first used for dependent gamma random variables. While the methodology works for an arbitrary number of dependent random densities, with each pair having their own dependent structure, the mathematics and estimation algorithm is focused on two dependent random density functions. Simulations and a real data example are presented.     

Swendsen-Wang Cuts sampling for spatially constrained Dirichlet process mixture models

Spatially constrained Dirichlet process mixture models are springing up in image processing in recent years. However, inference for the model is NP-hard. Gibbs sampling which is a generic Markov chain Monte Carlo technique is commonly employed for the model inference. It needs to traverse all the nodes of the constructed graph in each iteration. The sampling process hardly crosses over the intermediate low probabilistic state. In addition, it is not well informed by the spatial relationship in the sampling process. In this paper, a spatially dependent split-merge algorithm for sampling the MRF/DPMM model based on Swendsen-Wang Cuts is proposed. It is a state of the art algorithm which combines the spatial relationship to direct the sampling, and lessen the mixing time drastically. In this algorithm, a set of nodes are being frozen together according to the discriminative probability of the edges between neighboring nodes. The frozen nodes update their states simultaneously in contrast to the single node update in a Gibbs sampling. The final step of the algorithm is to accept the proposed new state according to the Metropolis Hasting scheme, in which only the ratio of posterior distribution needs to be calculated in each iteration. Experimental results demonstrated that the proposed sampling algorithm is able to reduce the mixing time considerably. At the same time, it can obtain comparably stable results with a random initial state.     

Dirichlet 过程及其在自然语言处理中的应用

Dirichlet过程是一种典型的变参数贝叶斯模型,其优点是参数的个数和性质灵活可变,可通过模型和数据来自主地计算,近年来它已成为机器学习和自然语言处理研究领域中的一个研究热点。该文较为系统的介绍了Dirichlet过程的产生、发展,并重点介绍了其模型计算,同时结合自然语言处理中的具体应用问题进行了详细分析。最后讨论了Dirichlet过程未来的研究方向和发展趋势。     

Design and implementation of S-MARKS: A secure middleware for pervasive computing applications

As portable devices have become a part of our everyday life, more people are unknowingly participating in a pervasive computing environment. People engage with not a single device for a specific purpose but many devices interacting with each other in the course of ordinary activity. With such prevalence of pervasive technology, the interaction between portable devices needs to be continuous and imperceptible to device users. Pervasive computing requires a small, scalable and robust network which relies heavily on the middleware to resolve communication and security issues. In this paper, we present the design and implementation of S-MARKS which incorporates device validation, resource discovery and a privacy module.     

A unified learning framework for object detection and classification using nested cascades of boosted classifiers

In this paper a unified learning framework for object detection and classification using nested cascades of boosted classifiers is proposed. The most interesting aspect of this framework is the integration of powerful learning capabilities together with effective training procedures, which allows building detection and classification systems with high accuracy, robustness, processing speed, and training speed. The proposed framework allows us to build state of the art face detection, eyes detection, and gender classification systems. The performance of these systems is validated and analyzed using standard face databases (BioID, FERET and CMU-MIT), and a new face database (UCHFACE).

A Dirichlet process mixture model for the analysis of correlated binary responses

The multivariate probit model is a popular choice for modelling correlated binary responses. It assumes an underlying multivariate normal distribution dichotomized to yield a binary response vector. Other choices for the latent distribution have been suggested, but basically all models assume homogeneity in the correlation structure across the subjects. When interest lies in the association structure, relaxing this homogeneity assumption could be useful. The latent multivariate normal model is replaced by a location and association mixture model defined by a Dirichlet process. Attention is paid to the parameterization of the covariance matrix in order to make the Bayesian computations convenient. The approach is illustrated on a simulated data set and applied to oral health data from the Signal Tandmobiel^® study to examine the hypothesis that caries is mainly a spatially local disease.     

Bayesian learning of finite generalized inverted Dirichlet mixtures: Application to object classification and forgery detection

The advent of mixture models has opened the possibility of flexible models which are practical to work with. A common assumption is that practitioners typically expect that data are generated from a Gaussian mixture. The inverted Dirichlet mixture has been shown to be a better alternative to the Gaussian mixture and to be of significant value in a variety of applications involving positive data. The inverted Dirichlet is, however, usually undesirable, since it forces an assumption of positive correlation. Our focus here is to develop a Bayesian alternative to both the Gaussian and the inverted Dirichlet mixtures when dealing with positive data. The alternative that we propose is based on the generalized inverted Dirichlet distribution which offers high flexibility and ease of use, as we show in this paper. Moreover, it has a more general covariance structure than the inverted Dirichlet. The proposed mixture model is subjected to a fully Bayesian analysis based on Markov Chain Monte Carlo (MCMC) simulation methods namely Gibbs sampling and Metropolis–Hastings used to compute the posterior distribution of the parameters, and on Bayesian information criterion (BIC) used for model selection. The adoption of this purely Bayesian learning choice is motivated by the fact that Bayesian inference allows to deal with uncertainty in a unified and consistent manner. We evaluate our approach on the basis of two challenging applications concerning object classification and forgery detection.     

Community detection from biological and social networks: A comparative analysis of metaheuristic algorithms

In order to analyze complex networks to find significant communities, several methods have been proposed in the literature. Modularity optimization is an interesting and valuable approach for detection of network communities in complex networks. Due to characteristics of the problem dealt with in this study, the exact solution methods consume much more time. Therefore, we propose six metaheuristic optimization algorithms, which each contain a modularity optimization approach. These algorithms are the original Bat Algorithm (BA), Gravitational Search Algorithm (GSA), modified Big Bang–Big Crunch algorithm (BB-BC), improved Bat Algorithm based on the Differential Evolutionary algorithm (BADE), effective Hyperheuristic Differential Search Algorithm (HDSA) and Scatter Search algorithm based on the Genetic Algorithm (SSGA). Four of these algorithms (HDSA, BADE, SSGA, BB-BC) contain new methods, whereas the remaining two algorithms (BA and GSA) use original methods. To clearly demonstrate the performance of the proposed algorithms when solving the problems, experimental studies were conducted using nine real-world complex networks − five of which are social networks and the rest of which are biological networks. The algorithms were compared in terms of statistical significance. According to the obtained test results, the HDSA proposed in this study is more efficient and competitive than the other algorithms that were tested.     

Oil spill detection from SAR intensity imagery using a marked point process

This paper presents a new algorithm for the detection of oil spill from SAR intensity images. The proposed algorithm combines the marked point process, Bayesian inference and Markov Chain Monte Carlo (MCMC) technique. In this paper, the candidates of oil spills or dark spots in a SAR intensity image are characterized by a Poisson marked point process. The marked point process is formed by a group of random points (as a point process modelling the locations of oil spills) and a set of parameters including geometric parameters of windows centred at the random points and gamma distribution parameters (as the marks attaching to each point). As a result, the candidates of oil spills are represented by a group of windows, in which the intensities of pixels follow independent and identical gamma distribution with lower mean than that for the identical gamma distribution of the pixels out of windows. Following the Bayesian paradigm, the posterior distribution, which characterizes the locations and statistical distributions of oil spills, can be obtained up to a normalizing constant. In order to simulate from the posterior distribution and to estimate the parameters of the posterior distribution, the Revisable Jump MCMC (RJMCMC) algorithm is used. The optimal locations and sizes of dark spots are obtained by a maximum a posteriori (MAP) algorithm. The proposed approach is tested using Radarsat-1 SAR images with oil spills indicated by human analysts. The results show that the proposed approach works well and is very promising.     

直拉硅单晶生长过程建模与控制研究综述

硅单晶是最重要的半导体材料,90%的半导体器件和集成电路芯片都制作在硅单晶上.随着集成电路技术的快速发展,对硅单晶的品质要求也不断提高.直拉法是生产硅单晶的主要方法,其科学原理与方法、生长技术与工艺、控制策略与手段一直是理论界和产业界高度关注和不断研究的热点.本文针对直拉法电子级硅单晶生长过程,以晶体生长基本原理为基础,从生长建模、变量检测、控制方法等方面进行了全面的阐述,特别针对当今大尺寸、高品质硅单晶生长的要求,总结了目前所取得的主要研究成果与面临的问题,并提出了相应的研究思路和方法.     

Quality detection and classification for ultrasonic welding of carbon fiber composites using time-series data and neural network methods

Ultrasonic welding is a novel and efficient technique for joining carbon fiber composites in the automotive industry. Weld quality detection and classification is important to its adoption and deep neural network models are a promising method for this purpose. However, it is difficult to collect the large volume of data needed to train these models with laboratory experiments due to the cost of the materials and cost of weld experiments. Using a limited set of experimental data, a copula multivariate Monte Carlo simulation is proposed to generate large data sets of time-series process signals with similar statistical distributions as the experimental data. The experimental data and simulated data are used to train Bayesian regularized neural network (BRNN) and convolutional neural network (CNN) models to predict weld quality classifications in ultrasonic welding. The results show that BRNN and CNN have similar classification accuracy. But CNN has an advantage in training efficiency compared with BRNN. Both neural-network-based methods were found to be more accurate than support vector machine and k-nearest neighbor methods, when applied to both features extracted from signals and full time-series-based process signals.