Accounting for regime and parameter uncertainty in regime-switching models

As investment guarantees become increasingly complex, realistic simulation of the price becomes more critical. Currently, regime-switching models are commonly used to simulate asset returns. Under a regime switching model, simulating random asset streams involves three steps: (i) estimate the model parameters given the number of regimes using maximum likelihood, (ii) choose the number of regimes using a model selection criteria, and (iii) simulate the streams using the optimal number of regimes and parameter values. This method, however, does not properly incorporate regime or parameter uncertainty into the generated asset streams and therefore into the price of the guarantee. To remedy this, this article adopts a Bayesian approach to properly account for those two sources of uncertainty and improve pricing.     

Refinements of two-sided bounds for renewal equations

Many quantities of interest in the study of renewal processes may be expressed as the solution to a special type of integral equation known as a renewal equation. The main purpose of this paper is to provide bounds for the solution of renewal equations based on various reliability classifications. Exponential and nonexponential types of inequalities are derived. In particular, two-sided bounds with specific reliability conditions become sharp. Finally, some examples including ultimate ruin for the classical Poisson model with time-dependent claim sizes, the joint distribution of the surplus prior to and at ruin, and the excess life time, are provided.     

A very fast algorithm for matrix factorization

We present a very fast algorithm for general matrix factorization of a data matrix for use in the statistical analysis of high-dimensional data via latent factors. Such data are prevalent across many application areas and generate an ever-increasing demand for methods of dimension reduction in order to undertake the statistical analysis of interest. Our algorithm uses a gradient-based approach which can be used with an arbitrary loss function provided the latter is differentiable. The speed and effectiveness of our algorithm for dimension reduction is demonstrated in the context of supervised classification of some real high-dimensional data sets from the bioinformatics literature.     

Worst case risk measurement: Back to the future?

This paper studies the problem of finding best-possible upper bounds on a rich class of risk measures, expressible as integrals with respect to measures, under incomplete probabilistic information. Both univariate and multivariate risk measurement problems are considered. The extremal probability distributions, generating the worst case scenarios, are also identified.The problem of worst case risk measurement has been studied extensively by Etienne De Vijlder and his co-authors, within the framework of finite-dimensional convex analysis. This paper revisits and extends some of their results.     

基于高光谱成像的纸质文物“狐斑”检测方法研究

受保存条件影响,很多纸质文物表面会形成狐斑（foxing）,如果不能进行有效监测和科学判断,会进而影响纸质文物安全。纸质文物狐斑病害检测存在滞后性、主观性等问题,对于书画藏品被墨色、颜料及印章等覆盖的区域更是难以通过肉眼进行识别,因此,基于文物的预防性保护理念,亟待开发对于狐斑高效、精确识别的无损检测技术。可见光-近红外高光谱图像结合了光谱和图像,包含丰富的空间信息与光谱信息,可以实现无损批量地平面采集样本光谱信息。该研究提出一种基于高光谱成像技术检测纸质文物狐斑的快速识别方法,获取模拟纸质文物在360～970 nm的高光谱图像,因360～450 nm受噪声影响过大,所以选择剔除这部分光谱数据;选取感兴趣区域并获取相应的平均光谱反射率,比较健康区域与被狐斑感染区域,发现两者的光谱曲线存在差异;在450～600 nm附近,受狐斑影响区域比健康区域的光谱反射率偏高,并在600 nm附近出现波峰形态;而在600～900 nm范围内,被感染区域与健康区域的光谱都趋于平稳,两者之间差异逐渐减小。选取从特征波长对应的图像中提取的特征信息建立图像识别模型,运用波段运算观察狐斑图像特征,狐斑的大小和分布情况都能清晰地显示,但与印章和墨迹重叠部分,狐斑被印章和墨迹遮盖,难以识别;利用最小噪声分离,虽然不同部分有重叠,但能发现仅凭肉眼难以识别的隐藏的狐斑;180条高光谱数据（450～970 nm）建立狐斑判别模型,随机地分为120条数据为训练集,60条数据为测试集,应用K-近邻法与BP神经网络建立纸质文物狐斑光谱判别模型,总体上两种方法对狐斑判别率分别达到73.3%和85%;BP神经网络相较于K-近邻模型,总体判别率更高,识别效果也更好。结果表明,利用高光谱成像可高效准确识别纸质文物狐斑,为后续研究狐斑分布发展提供可靠的技术手段,也为博物馆馆藏文物的保存提供指导意见。     

基于Mask RCNN的桥梁裂缝检测方法设计及研究北大核心CSCD

裂缝是桥梁道路上常见的一种病害,针对其检测准确率有待提高的问题,提出了基于Mask RCNN(region-based convolutional neural networks)的桥梁裂缝检测算法,设计了语义增强模块(semantic enhancement module,SEM),将该模块与特征金字塔网络(feature pyramid network,FPN)相结合,通过特征融合Add计算得到新的多尺度特征图feature maps。针对裂缝形态复杂多样存在识别困难的问题,将裂缝做了两类划分进行检测,并制定了两种策略进行对比实验。实验结果表明:该文中改进的方法可以得到更好的检测结果,检测准确率Accuracy可达99.8%,平均检测精度(mean average precision,mAP)提高了12.6%。     

Moment-based modulation classification for Rician time-varying aeronautical channels

Modulation classification is a classical topic in the field of signal classification, and is of great significance in various applications including aeronautical communications. However, changes in the features of the signals transmitted through the Rician time-varying channels deteriorate the performance of modulation classification. In this study, a robust moment-based algorithm is proposed to counter the influence of the channels. The statistical correlation of the received samples are utilized for the estimation on the Rician factor. According to the derivation of the relationship between the moments of the transmitted samples and the received samples, the features utilized for classification are compensated mathematically. Finally, decision tree and random forest classifier are presented for classification. The estimation and classification performance of the proposed algorithm are analyzed for aeronautical en-route (EN) and arrival/takeoff (AT) channels through simulations. Meanwhile, real world signals of L-band digital aeronautical communication system (L-DACS1) are utilized for verification. The results show that our algorithm has high classification accuracy and low computational complexity.     

Accurate modulation classification under impaired wireless channels via shallow convolutional neural networks

Classifying the modulation type of radio signals plays an important role in current and future wireless communication systems. We present a modulation classification method based on convolutional neural networks that reaches high accuracy in face of various channel characteristics and signal conditions without requiring the network to have a very large depth. Experiment results show that the proposed method reaches accurate classification under different system impairment settings that include sampling rate offset, carrier frequency offset, multi-path fading, and additive white Gaussian noise. For instance, compared to a state-of-the-art method, accuracy is improved up to 25% in classifying difficult modulation types under system impairments. Source code of the proposed method is available online.     

Eigenimage-Based Facial Recognition Technique Using Gradient Covariance

In this paper, we propose a novel eigenface-based face recognition approach. First we describe a continuous model for facial feature extraction, involving the covariance operator associated with gradient of the 2-D image function. Then we provide a discretized version of this model. Face identification and verification procedures, using a supervised classification technique, are also proposed.