Evaluating the Efficacy of Latent Variables in Mitigating Data Poisoning Attacks in the Context of Bayesian Networks:An Empirical Study

Bayesian networks are a powerful class of graphical decision models used to represent causal relationships among variables.However,the reliability and integrity of learned Bayesian network models are highly dependent on the quality of incoming data streams.One of the primary challenges with Bayesian networks is their vulnerability to adversarial data poisoning attacks,wherein malicious data is injected into the training dataset to negatively influence the Bayesian network models and impair their performance.In this research paper,we propose an efficient framework for detecting data poisoning attacks against Bayesian network structure learning algorithms.Our framework utilizes latent variables to quantify the amount of belief between every two nodes in each causal model over time.We use our innovative methodology to tackle an important issue with data poisoning assaults in the context of Bayesian networks.With regard to four different forms of data poisoning attacks,we specifically aim to strengthen the security and dependability of Bayesian network structure learning techniques,such as the PC algorithm.By doing this,we explore the complexity of this area and offer workablemethods for identifying and reducing these sneaky dangers.Additionally,our research investigates one particular use case,the“Visit to Asia Network.”The practical consequences of using uncertainty as a way to spot cases of data poisoning are explored in this inquiry,which is of utmost relevance.Our results demonstrate the promising efficacy of latent variables in detecting and mitigating the threat of data poisoning attacks.Additionally,our proposed latent-based framework proves to be sensitive in detecting malicious data poisoning attacks in the context of stream data.

Discovering Latent Variables for the Tasks With Confounders in Multi-Agent Reinforcement Learning

Efficient exploration in complex coordination tasks has been considered a challenging problem in multi-agent reinforcement learning(MARL). It is significantly more difficult for those tasks with latent variables that agents cannot directly observe. However, most of the existing latent variable discovery methods lack a clear representation of latent variables and an effective evaluation of the influence of latent variables on the agent. In this paper, we propose a new MARL algorithm based on the soft actor-critic method for complex continuous control tasks with confounders. It is called the multi-agent soft actor-critic with latent variable(MASAC-LV) algorithm, which uses variational inference theory to infer the compact latent variables representation space from a large amount of offline experience.Besides, we derive the counterfactual policy whose input has no latent variables and quantify the difference between the actual policy and the counterfactual policy via a distance function. This quantified difference is considered an intrinsic motivation that gives additional rewards based on how much the latent variable affects each agent. The proposed algorithm is evaluated on two collaboration tasks with confounders, and the experimental results demonstrate the effectiveness of MASAC-LV compared to other baseline algorithms.

Pyramid-VAE-GAN:Transferring hierarchical latent variables for image inpainting

Significant progress has been made in image inpainting methods in recent years.However,they are incapable of producing inpainting results with reasonable structures,rich detail,and sharpness at the same time.In this paper,we propose the Pyramid-VAE-GAN network for image inpainting to address this limitation.Our network is built on a variational autoencoder(VAE)backbone that encodes high-level latent variables to represent complicated high-dimensional prior distributions of images.The prior assists in reconstructing reasonable structures when inpainting.We also adopt a pyramid structure in our model to maintain rich detail in low-level latent variables.To avoid the usual incompatibility of requiring both reasonable structures and rich detail,we propose a novel cross-layer latent variable transfer module.This transfers information about long-range structures contained in high-level latent variables to low-level latent variables representing more detailed information.We further use adversarial training to select the most reasonable results and to improve the sharpness of the images.Extensive experimental results on multiple datasets demonstrate the superiority of our method.Our code is available at https://github.com/thy960112/Pyramid-VAE-GAN.

A semantic and emotion-based dual latent variable generation model for a dialogue system

With the development of intelligent agents pursuing humanisation,artificial intelligence must consider emotion,the most basic spiritual need in human interaction.Traditional emotional dialogue systems usually use an external emotional dictionary to select appropriate emotional words to add to the response or concatenate emotional tags and semantic features in the decoding step to generate appropriate responses.However,selecting emotional words from a fixed emotional dictionary may result in loss of the diversity and consistency of the response.We propose a semantic and emotion-based dual latent variable generation model(Dual-LVG)for dialogue systems,which is able to generate appropriate emotional responses without an emotional dictionary.Different from previous work,the conditional variational autoencoder(CVAE)adopts the standard transformer structure.Then,Dual-LVG regularises the CVAE latent space by introducing a dual latent space of semantics and emotion.The content diversity and emotional accuracy of the generated responses are improved by learning emotion and semantic features respectively.Moreover,the average attention mechanism is adopted to better extract semantic features at the sequence level,and the semi-supervised attention mechanism is used in the decoding step to strengthen the fusion of emotional features of the model.Experimental results show that Dual-LVG can successfully achieve the effect of generating different content by controlling emotional factors.

Variability in Latent Heat Flux over the Tropical Pacific in Association with Recent Two ENSO Events

This paper analyzed the variations of latent heat flux (LHF) over the tropical Pacific in the period 1978-1988 by using COADS (Comprehensive Ocean and Atmospheric Data Set). It has been founded that the interannual variabili ty of LHF exhibits strong ENSO signal, with the significant increasing LHF during the recent two warm events, i.e., 1982 / 83 and 1986 / 87 and decreasing LHF in the cold episodes. However the longitudinal distribution of the LHF departures varies from event to event. In the eastern Pacific, the specific humidity difference at air-sea interface (qs -qa) makes a dominant contribution to the interannual variability of LHF ( r = 0.73 ), while in the western Pacific the surface wind speed, W and the qs - qa make nearly equal contribution to that of LHF.

Parking choice behaviour analysis of rural residents based on the latent variable random forest model

The imbalance of rural parking supply and demand has a great impact on traffic congestion and environmental pollution,which has attracted the attention of many scholars as well as policymakers.However,most of the current research on parking choice focuses on urban business and residential areas rather than on rural parking choice behaviour,and focuses on the analysis of observable factors,ignoring the internal relationship with potential variables.This study considers the heterogeneity of individuals and uses the random forest(RF)algorithm to construct a model of rural residents’willingness to choose parking with both latent and explicit variables,to explore how much and in what ways individual characteristics and parking characteristics affect rural residents’parking choices,and to explore parking planning programmes and strategies that are truly applicable to rural areas.The results of the study suggest that the safety and convenience of the parking environment are key factors influencing the parking choice behaviour of rural residents,and can greatly improve the predictive accuracy of the parking willingness model.Upon comparison,it is found that the application of the RF algorithm is also significantly better than the logit model in terms of prediction effect,indicating that there is a nonlinear effect among the factors influencing the parking choice behaviour of rural residents and that the RF model with the addition of latent variables provides a better explanatory ability for the study of the parking choice behaviour of rural residents.

Temporally Preserving Latent Variable Models:Offline and Online Training for Reconstruction and Interpretation of Fault Data for Gearbox Condition Monitoring

Latent variable models can effectively determine the condition of essential rotating machinery without needing labeled data.These models analyze vibration data via an unsupervised learning strategy.Temporal preservation is necessary to obtain an informative latent manifold for the fault diagnosis task.In a temporalpreserving context,two approaches exist to develop a condition-monitoring methodology:offline and online.For latent variable models,the available training modes are not different.While many traditional methods use offline training,online training can dynamically adjust the latent manifold,possibly leading to better fault signature extraction from the vibration data.This study explores online training using temporal-preserving latent variable models.Within online training,there are two main methods:one focuses on reconstructing data and the other on interpreting the data components.Both are considered to evaluate how they diagnose faults over time.Using two experimental datasets,the study confirms that models from both training modes can detect changes in machinery health and identify faults even under varying conditions.Importantly,the complementarity of offline and online models is emphasized,reassuring their versatility in fault diagnostics.Understanding the implications of the training approach and the available model formulations is crucial for further research in latent variable modelbased fault diagnostics.

Relationship Between Alcohol Drinking and Alcohol-related Health Problems

Objective To study the relationship between drinking environment, attitudes and situation and alcohol-related health problems. Methods A sample of 2327 respondents was randomly collected from Wuhan, Hubei Province in China by a face-to-face interview. The structural equation modeling analysis was performed for the data collected. Results Both parents' drinking behaviors and respondents' drinking situation strongly impacted the alcohol-related problems and diseases. Friends' or peers' drinking behaviors influenced the respondents' drinking attitudes and behaviors. Males experienced more alcohol-related problems and diseases than females. Conclusions Comparatively, parents' drinking behaviors exert the most significant influence on drinkers. Therefore, it is beneficial to restrict parents' drinking behaviors for the offsprings and the whole society, and an intensive professional education in early motherhood is also necessary for Chinese women.

A novel deep learning framework with variational auto-encoder for indoor air quality prediction

Exposure to poor indoor air conditions poses significant risks to human health, increasing morbidity and mortality rates. Soft measurement modeling is suitable for stable and accurate monitoring of air pollutants and improving air quality. Based on partial least squares (PLS), we propose an indoor air quality prediction model that utilizes variational auto-encoder regression (VAER) algorithm. To reduce the negative effects of noise, latent variables in the original data are extracted by PLS in the first step. Then, the extracted variables are used as inputs to VAER, which improve the accuracy and robustness of the model. Through comparative analysis with traditional methods, we demonstrate the superior performance of our PLS-VAER model, which exhibits improved prediction performance and stability. The root mean square error (RMSE) of PLS-VAER is reduced by 14.71%, 26.47%, and 12.50% compared to single VAER, PLS-SVR, and PLS-ANN, respectively. Additionally, the coefficient of determination (R2) of PLS-VAER improves by 13.70%, 30.09%, and 11.25% compared to single VAER, PLS-SVR, and PLS-ANN, respectively. This research offers an innovative and environmentally-friendly approach to monitor and improve indoor air quality.

Examining unobserved factors associated with red light running in Vietnam:A latent class model analysis

Red-light running(RLR)is a crucial violation that causes traffic accidents and injuries.Understanding factors that affect RLR is very significant to reduce the potential of this violation.Current studies have paid considerable attention to the observable factors,but not to unobservable factors.This study aims to examine the effects of observable and unobservable factors on RLR.This study uses a latent class model(LCM)to assign individuals into two classes—red-light-respectful and red-light-disrespectful road users—by surveying 751 respondents who use private transportation modes.This study incorporates psychological determinants into the LCM to account for unobservable factors.The contribution of this study is the in-depth investigation into law-respectful and law-disrespectful behaviours and intentional and unintentional violators.Such a study has not yet been conducted in the existing literature.In addition,a comprehensive comparison of the LCM and a traditional ordered probit model was conducted.Overall,the results suggest that the LCM is superior to the model that does not consider latent classes.Our estimation results are in alignment with previous studies on RLR:males,younger drivers/riders,less educated road users and motorcyclists are more likely to run red lights.An analysis of the latent variables shows that surrounding conditions—the behaviour of other violators,the absence of traffic police,and long waiting times—increase the possibility of violations.Based on these results,we provide suggestions to policymakers and traffic engineers:the implementation of enforcement cameras and penalties for violators are critical countermeasures to minimize the potential of RLR.

A linear varying coefficient ARCH-M model with a latent variable

Motivated by the psychological factor of time-varying risk-return relationship, this paper studies a linear varying coefficient ARCH-M model with a latent variable. Due to the unobservable property of the latent variable, a corrected likelihood method is employed for parametric estimation. Estimators are proved to be consistent and asymptotically normal under certain regularity conditions. A simple test statistic is also proposed for testing latent variable effect. Simulation results confirm that the proposed estimators and test perform well.The model is further applied to examine whether the risk-return relationship depends on investor's sentiment in American Market and some explainable results are obtained.

Evaluation of the Conjoint Efficacy in Chinese Medicine with the Longitudinal Latent Variable Linear Mixed Model

Chinese medicine （CM） clinical efficacy evaluation research involves the longitudinal multivariate measurement which means that patients are measured repeatedly and each patient is measured by several indicators on each fixed cross-section. Although each indicator can be evaluated separately with a longitudinal linear mixed model, it is important to consider all the endpoints together especially when researchers pay special attention to the change of the conjoint efficacy for several indicators in one patient. In this article, we introduce a latent variable linear mixed model to the CM conjoint efficacy evaluation and discuss why and how to analyze the longitudinal multivariate endpoint data in the clinical CM efficacy evaluation research. It may lead to the new insight of using such methodology in the field of conjoint efficacy evaluating of CM study. And with the definition of syndrome and symptom in the CM theory, the applied discussion brings the insight of CM syndrome evaluating in future. We illustrate this methodology using an example of CM efficacy evaluation from an ischemic stroke research.

Dimensionality reduction with latent variable model

Over the past few decades, latent variable model （LVM）-based algorithms have attracted consid- erable attention for the purpose of data diInensional- ity reduction, which plays an important role in machine learning, pattern recognition, and computer vision. LVM is an effective tool for modeling density of the observed data. It has been used in dimensionality reduction for dealing with the sparse observed samples. In this paper, two LVM-based dimensionality reduction algorithms are presented firstly, i.e., supervised Gaussian process la- tent variable model and senti-supervised Gaussian pro- cess latent variable model. Then, we propose an LVM- based transfer learning model to cope with the case that samples are not independent identically distributed. In the end of each part, experimental results are given to demonstrate the validity of the proposed dimensionality reduction algorithms.

Tolerance Limits Under Gamma Mixtures:Application in Hydrology

In this study,the authors proposed upper tolerance limits for the gamma mixture distribution based on generalized fiducial inference,and an MCMC simulation is performed to sample from the generalized fiducial distributions.The simulation results and a real hydrological data example show that the proposed tolerance limits are more efficient.

Quality-related locally weighted soft sensing for non-

Soft sensors are widely used to predict quality variables which are usually hard to measure.It is necessary to construct an adaptive model to cope with process non-stationaries.In this study,a novel quality-related locally weighted soft sensing method is designed for non-stationary processes based on a Bayesian network with latent variables.Specifically,a supervised Bayesian network is proposed where quality-oriented latent variables are extracted and further applied to a double-layer similarity meas-urement algorithm.The proposed soft sensing method tries to find a general approach for non-stationary processes via quality-related information where the concepts of local similarities and window confidence are explained in detail.The performance of the developed method is demonstrated by application to a numerical example and a debutanizer column.It is shown that the proposed method outperforms competitive methods in terms of the accuracy of predicting key quality variables.

Semiparametric Analysis of Longitudinal Data with Informative Observation Times

In many longitudinal studies, observation times as well as censoring times may be correlated with longitudinal responses. This paper considers a multiplicative random effects model for the longitudinal response where these correlations may exist and a joint modeling approach is proposed via a shared latent variable. For inference about regression parameters, estimating equation approaches are developed and asymptotic properties of the proposed estimators are established. The finite sample behavior of the methods is examined through simulation studies and an application to a data set from a bladder cancer study is provided for illustration.

Statistical inference for zero-and-one-inflated poisson models

In this paper, a zero-and-one-inflated Poisson (ZOIP) model is studied. The maximum likelihoodestimation and the Bayesian estimation of the model parameters are obtained based on dataaugmentation method. A simulation study based on proposed sampling algorithm is conductedto assess the performance of the proposed estimation for various sample sizes. Finally, two realdata-sets are analysed to illustrate the practicability of the proposed method.

Bayesian Empirical Likelihood Estimation of Quantile Structural Equation Models

Structural equation model(SEM) is a multivariate analysis tool that has been widely applied to many fields such as biomedical and social sciences. In the traditional SEM, it is often assumed that random errors and explanatory latent variables follow the normal distribution, and the effect of explanatory latent variables on outcomes can be formulated by a mean regression-type structural equation. But this SEM may be inappropriate in some cases where random errors or latent variables are highly nonnormal. The authors develop a new SEM, called as quantile SEM(QSEM), by allowing for a quantile regression-type structural equation and without distribution assumption of random errors and latent variables. A Bayesian empirical likelihood(BEL) method is developed to simultaneously estimate parameters and latent variables based on the estimating equation method. A hybrid algorithm combining the Gibbs sampler and Metropolis-Hastings algorithm is presented to sample observations required for statistical inference. Latent variables are imputed by the estimated density function and the linear interpolation method. A simulation study and an example are presented to investigate the performance of the proposed methodologies.

Option Pricing under the Double Exponential Jump-Diffusion Model with Stochastic Volatility and Interest Rate

This paper proposes an efficient option pricing model that incorporates stochastic interest rate(SIR),stochastic volatility(SV),and double exponential jump into the jump-diffusion settings.The model comprehensively considers the leptokurtosis and heteroscedasticity of the underlying asset’s returns,rare events,and an SIR.Using the model,we deduce the pricing characteristic function and pricing formula of a European option.Then,we develop the Markov chain Monte Carlo method with latent variable to solve the problem of parameter estimation under the double exponential jump-diffusion model with SIR and SV.For verification purposes,we conduct time efficiency analysis,goodness of fit analysis,and jump/drift term analysis of the proposed model.In addition,we compare the pricing accuracy of the proposed model with those of the Black-Scholes and the Kou(2002)models.The empirical results show that the proposed option pricing model has high time efficiency,and the goodness of fit and pricing accuracy are significantly higher than those of the other two models.

Dirichlet process and its developments: a survey

The core of the nonparametric/semiparametric Bayesian analysis is to relax the particular parametric assumptions on the distributions of interest to be unknown and random,and assign them a prior.Selecting a suitable prior therefore is especially critical in the nonparametric Bayesian fitting.As the distribution of distribution,Dirichlet process(DP)is the most appreciated nonparametric prior due to its nice theoretical proprieties,modeling flexibility and computational feasibility.In this paper,we review and summarize some developments of DP during the past decades.Our focus is mainly concentrated upon its theoretical properties,various extensions,statistical modeling and applications to the latent variable models.