Latent variable model

About: Latent variable model is a research topic. Over the lifetime, 3589 publications have been published within this topic receiving 235061 citations.

Sequential Latent Knowledge Selection for Knowledge-Grounded Dialogue

Abstract: Knowledge-grounded dialogue is a task of generating an informative response based on both discourse context and external knowledge. As we focus on better modeling the knowledge selection in the multi-turn knowledge-grounded dialogue, we propose a sequential latent variable model as the first approach to this matter. The model named sequential knowledge transformer (SKT) can keep track of the prior and posterior distribution over knowledge; as a result, it can not only reduce the ambiguity caused from the diversity in knowledge selection of conversation but also better leverage the response information for proper choice of knowledge. Our experimental results show that the proposed model improves the knowledge selection accuracy and subsequently the performance of utterance generation. We achieve the new state-of-the-art performance on Wizard of Wikipedia (Dinan et al., 2019) as one of the most large-scale and challenging benchmarks. We further validate the effectiveness of our model over existing conversation methods in another knowledge-based dialogue Holl-E dataset (Moghe et al., 2018).

Modeling Incomplete Longitudinal Substance Use Data Using Latent Variable Growth Curve Methodology

Abstract: Longitudinal data sets typically suffer from attrition and other forms of missing data. When this common problem occurs, several researchers have demonstrated that correct maximum likelihood estimation with missing data can be obtained under mild assumptions concerning the missing data mechanism. With reasonable substantive theory, a mixture of cross-sectional and longitudinal methods developed within multiple-group structural equation modeling can provide a strong basis for inference about developmental change. Using an approach to the analysis of missing data, the present study investigated developmental trends in adolescent (N = 759) alcohol, marijuana, and cigarette use across a 5-year period using multiple-group latent growth modeling. An associative model revealed that common developmental trends existed for all three substances. Age and gender were included in the model as predictors of initial status and developmental change. Findings discuss the utility of latent variable structural equation modeling techniques and missing data approaches in the study of developmental change.

INLR, implicit non-linear latent variable regression

Abstract: A simple way to develop non-linear PLS models is presented, INLR (implicit non-linear latent variable regression). The paper shows that by simply added squared x-variables x2a, both the square and cross terms of the latent variables are implicitly included in the resulting PLS model. This approach works when X itself is well modelled by a projection model T*PT. Hence, if a latent structure is present in X, it is not necessary to include the cross terms of the X-variables in the polynomial expansion. Analogously, with cubic non-linearities, expanding X with cubic terms x3a is sufficient. INLR is attractive in that all essential features of PLS are preserved i.e. (a) it can handle many noisy and collinear variables, (b) it is stable and gives reliable results and (c) all PLS plots and diagnostics still apply. The principles of INLR are outlined and illustrated with three chemical examples where INLR improved the modelling and predictions compared with ordinary linear PLS. © 1997 John Wiley & Sons, Ltd.

Interactions of latent variables in structural equation models

Abstract: Interactions of variables occur in a variety of statistical analyses The best known procedures for models with interactions of latent variables are technically demanding Not only does the potential user need to be familiar with structural equation modeling (SEM), but the researcher must be familiar with programming nonlinear and linear constraints and must be comfortable with fairly large and complicated models This article provides a largely nontechnical description of an alternative two‐stage least squares (2SLS) technique to include interactions of latent variables in SEM The method requires the selection of instrumental variables and we give rules for their selection in the most common cases We compare the 2SLS method to the alternatives Some of the important advantages of the 2SLS are that it can handle nonnormal observed variables, is readily available in major statistical software packages, and has a known asymptotic distribution In providing the comparisons, we reanalyze all the interaction

Latent variable models for longitudinal data with multiple continuous outcomes.

Abstract: Multiple outcomes are often used to properly characterize an effect of interest. This paper proposes a latent variable model for the situation where repeated measures over time are obtained on each outcome. These outcomes are assumed to measure an underlying quantity of main interest from different perspectives. We relate the observed outcomes using regression models to a latent variable, which is then modeled as a function of covariates by a separate regression model. Random effects are used to model the correlation due to repeated measures of the observed outcomes and the latent variable. An EM algorithm is developed to obtain maximum likelihood estimates of model parameters. Unit-specific predictions of the latent variables are also calculated. This method is illustrated using data from a national panel study on changes in methadone treatment practices.