Data-driven preventive maintenance for a heterogeneous machine portfolio

We describe a data-driven approach to optimize periodic maintenance policies for a heterogeneous portfolio with different machine profiles. When insufficient data are available per profile to assess failure intensities and costs accurately, we pool the data of all machine profiles and evaluate the effect of (observable) machine characteristics by calibrating appropriate statistical models. This reduces maintenance costs compared to a stratified approach that splits the data into subsets per profile and a uniform approach that treats all profiles the same.     

Gompertz zero-inflated cure rate regression models applied to credit risk data

In the financial market, it is important to consider that there is a proportion of customers that have settled their debt in time zero, immediately recovering their ability to pay. In this context, in this paper, we propose a survival analysis methodology that allows the insertion of times equal to zero in scenarios where credit risk is observed. The proposed model addresses the survival analysis model of the zero-inflated cure rate which incorporates the heterogeneity of three subgroups (individuals having events in the initial time, and individuals not susceptible and susceptible to the event). In our proposal, all available survival data of customers are modeled considering that the number of competitive causes follows a Poisson distribution and the baseline risk function follows a Gompertz distribution. The model parameter estimation is obtained by the maximum likelihood estimation procedure and simulation studies are conducted to evaluate the estimators' performance. The studied methodology will be applied to a credit database provided by a financial institution in Brazil.     

Quantum Bayesian computation

Quantum Bayesian computation is an emerging field that levers the computational gains available from quantum computers. They promise to provide an exponential speed-up in Bayesian computation. Our article adds to the literature in three ways. First, we describe how quantum von Neumann measurement provides quantum versions of popular machine learning algorithms such as Markov chain Monte Carlo and deep learning that are fundamental to Bayesian learning. Second, we describe quantum data encoding methods needed to implement quantum machine learning including the counterparts to traditional feature extraction and kernel embeddings methods. Third, we show how quantum algorithms naturally calculate Bayesian quantities of interest such as posterior distributions and marginal likelihoods. Our goal then is to show how quantum algorithms solve statistical machine learning problems. On the theoretical side, we provide quantum versions of high dimensional regression, Gaussian processes and stochastic gradient descent. On the empirical side, we apply a quantum FFT algorithm to Chicago house price data. Finally, we conclude with directions for future research.     

A Bayesian record linkage model incorporating relational data

In this article, we introduce a novel Bayesian approach for linking multiple social networks in order to discover the same real world person having different accounts across networks. In particular, we develop a latent model that allows us to jointly characterize the network and linkage structures relying on both relational and profile data. In contrast to other existing approaches in the machine learning literature, our Bayesian implementation naturally provides uncertainty quantification via posterior probabilities for the linkage structure itself or any function of it. Our findings clearly suggest that our methodology can produce accurate point estimates of the linkage structure even in the absence of profile information, and also, in an identity resolution setting, our results confirm that including relational data into the matching process improves the linkage accuracy. We illustrate our methodology using real data from popular social networks such as Twitter , Facebook , and YouTube .     

A review of Bayesian dynamic forecasting models: Applications in marketing

We briefly review the main developments of Bayesian dynamic models. The emphasis is on marketing applications. Typical examples in this area are discussed. The concepts of monitoring and intervention are carefully explained with illustrative examples and open source computational routines. We avoid algebraic developments and instead use graphical examples to illustrate theoretical aspects. Two real-world problems using Bayesian dynamic models are discussed. Finally, we describe recent developments and alternative proposals to formally address the dependence when dealing with the modeling of multiple time series.     

Ill-posed problems in rheology

Experimental data are always noisy and often incomplete. This leads to ambiguities if one wants to infer from the data some functions, which are related to the measured quantity through an integral equation of the first kind. In rheology many of such so-called ill-posed problems appear. Two techniques to treat such problems, the regularization method and the maximum entropy method, are applied to the determination of the relaxation spectrum from data of small oscillatory shear flow. With simulated data from a reference spectrum it is discussed how the inferred spectrum depends on the region, in which data are available. It turns out that information about the asymptotic behavior of the measured quantity can be of great help in determining the full spectrum also from incomplete data.Dedicated to Prof. Dr. J. Meissner on the occasion of his 60th birthday.     

异质传感器数据的最优线性融合

研究了异质传感器对参数矢量进行的测量，测量数据基于线性均方估计的最优融合算法。提出和证明了异质传感器数据的最优线性融合定理，并得出“精度再差的传感器参与数据融合后都有利于提高系统的测量精度”这一结论。     

激波风洞信号测量仪

本文叙述了一种用于激波风洞实验研究中的以微机为基础的多功能智能测量仪表,叙述了其构成部分和性能、硬件结构和原理、软件编制以及将BASIC程序写入EPROM的过程。     

改进遗传人工神经网络在组合导航中的应用

鉴于常规卡尔曼滤波算法组合导航系统数据融合算法中,存在易于发散的缺陷,尝试将遗传优化人工神经网络引入组合导航系统中.针对传统遗传算法存在的易早熟、算法稳定性差、固定的交叉和变异概率影响收敛效果等缺点,采用浮点式编码方式,两两竞争的选择策略、引入突变操作、重新定义交叉算子和自适应的交叉变异算子等措施进行了遗传算法的改进.仿真结果表明,改进后的算法更为有效,并且精度与常规卡尔曼滤波算法相当.     

An Approach to NMR Assignment of Intrinsically Disordered Proteins

An efficient approach to NMR assignments in intrinsically disordered proteins is presented, making use of the good dispersion of cross peaks observed in [¹⁵N,¹³C′]‐ and [¹³C′,¹H^N]‐correlation spectra. The method involves the simultaneous collection of {3D (H)NCO(CAN)H and 3D (HACA)CON(CA)HA} spectra for backbone assignments via sequential H^N and H^α correlations and {3D (H)NCO(CACS)HS and 3D (HS)CS(CA)CO(N)H} spectra for side‐chain ¹H and ¹³C assignments, employing sequential ¹H data acquisitions with direct detection of both the amide and aliphatic protons. The efficacy of the approach for obtaining resonance assignments with complete backbone and side‐chain chemical shifts is demonstrated experimentally for the 61‐residue [¹³C,¹⁵N]‐labelled peptide of a voltage‐gated potassium channel protein of the Kv1.4 channel subunit. The general applicability of the approach for the characterisation of moderately sized globular proteins is also demonstrated.