Discrete Mixture Research Articles

Sub-Nanosheet Induced Inverse Growth of Negative Valency Au Clusters for Tumor Treatment by Enhanced Oxidative Stress.

Cluster aggregation states are thermodynamically favored at the subnanoscale, for which an inverse growth from nanoparticles to clusters may be realized on subnanometer supports. Herein, we develop Au-polyoxometalate-layered double hydroxide (Au-POM-LDH) sub-1 nm nanosheets (Sub-APL) based on the above strategy, where sub-1 nm Au clusters with negative valence are generated by the in situ disintegration of Au nanoparticles on POM-LDH supports. Sub-1 nm Au clusters with ultrahigh surface atom ratios exhibit remarkable efficiency for glutathione (GSH) depletion. The closely connected sub-1 nm Au with negative valence and POM hetero-units can promote the separation of hole-electrons, resulting in the enhanced reactive oxygen species (ROS) generation under ultrasound (US). Besides, the reversible redox of Mo in POM is able to deplete GSH and trigger chemodynamic therapy (CDT) simultaneously, further enhancing the oxidative stress. Consequently, the Sub-APL present 2-fold ROS generation under US and 7-fold GSH depletion compared to the discrete Au and POM-LDH mixture. Therefore, the serious imbalance of redox in the TME caused by the sharp increase of ROS and rapid decrease of GSH leads to death of tumor ultimately.

Monitoring mean of INAR(1) process with discrete mixture exponential innovations

Monitoring mean of INAR(1) process with discrete mixture exponential innovations

Reinforcement learning-based order-dispatching optimization in the ride-sourcing service

Since the emergence of ride-sourcing, the dispatching problem has been a hot research issue in the travel market. Previous studies designed various refined algorithms to improve the dispatching efficiency. However, the heterogeneity of passenger cancellation behavior and driver's work patterns is ignored in the modeling of these studies, which may result in the algorithms failing to entirely capture the randomness in the real world, thereby affecting dispatching efficiency. To fill this research gap, this paper employs a discrete choice model and Gaussian mixture model to model passenger cancellation behavior and extract driver working patterns, respectively. Furthermore, a cooperative deep deterministic policy gradient (DDPG) algorithm is proposed for dispatching. The algorithm constructs decision networks separately for drivers with different working patterns to learn the optimal dispatching decision and designs a centralized neural network to coordinate the dispatching of all vehicles to maximize the global benefit. A price incentive strategy is proposed to improve passenger loyalty by providing incentive prices. Moreover, based on the constructed discrete choice model, a simulation environment is designed to describe the cancellation behavior in the real world and support the training of the algorithm. Experimental results based on real data in Beijing show that the cooperative DDPG algorithm can improve the order response rate by 2% to 6% and platform revenue by 3% to 9% in multiple travel periods while providing a better travel experience to passengers. Additionally, the price incentive strategy also effectively reduces passenger cancellation.

Disease activity trajectories from childhood to adulthood in the population-based Nordic juvenile idiopathic arthritis cohort

ObjectivesTo identify long-term disease activity trajectories from childhood to adulthood by using the clinical Juvenile Arthritis Disease Activity Score (cJADAS10) in juvenile idiopathic arthritis (JIA). Second, to evaluate the contribution...

A Bayesian fine-mapping model using a continuous global-local shrinkage prior with applications in prostate cancer analysis

The aim of fine mapping is to identify genetic variants causally contributing to complex traits or diseases. Existing fine-mapping methods employ Bayesian discrete mixture priors and depend on a pre-specified maximum number of causal variants, which may lead to sub-optimal solutions. In this work, we propose a Bayesian fine-mapping method called h2-D2, utilizing a continuous global-local shrinkage prior. We also present an approach to define credible sets of causal variants in continuous prior settings. Simulation studies demonstrate that h2-D2 outperforms current state-of-the-art fine-mapping methods such as SuSiE and FINEMAP in accurately identifying causal variants and estimating their effect sizes. We further applied h2-D2 to prostate cancer analysis and discovered some previously unknown causal variants. In addition, we inferred 369 target genes associated with the detected causal variants and several pathways that were significantly over-represented by these genes, shedding light on their potential roles in prostate cancer development and progression.

Deinterleaving of Discrete Renewal Process Mixtures With Application to Electronic Support Measures

Deinterleaving of Discrete Renewal Process Mixtures With Application to Electronic Support Measures

On Discrete Mixture of Moment Exponential Using Lagrangian Probability Model: Properties and Applications in Count Data with Excess Zeros

On Discrete Mixture of Moment Exponential Using Lagrangian Probability Model: Properties and Applications in Count Data with Excess Zeros

Developmental trajectories of speech and language in neglected children aged 3 to 5 years: Results of the ELLAN study

BackgroundNeglected children are at high risk for significant difficulties in speech and language development. Because no longitudinal study has been conducted to date, the dynamic description of development during the preschool period is unknown. ObjectivesEstablish the developmental trajectories of speech sounds, receptive and expressive vocabulary, and morphosyntax among neglected children during the preschool years and compare them with those of non-neglected children. Participants and settingParticipants are 69 neglected children and 99 same age non-neglected peers (37 and 46 males respectively) recruited at 36 months of age. Data were collected at home. MethodsData were collected at six-month intervals between the ages of 3 and 5.5 years using psychometrically robust tools. Neglected and control groups were compared according to age using repeated measures ANOVAs on all variables. A discrete mixture model for clustering longitudinal data was used for testing the heterogeneity of the language trajectories among neglected children. ResultsThe language development of the neglected children as a whole group is lower than that of the control group for all variables. Two subgroups are identified within the neglected group: one with a developmental trajectory similar to that of the non-neglected children, and another whose trajectory is far below that of the control group. The effect sizes of these differences vary between 1.4 and 3 standard deviations under the mean. ConclusionsA large proportion of neglected children present significant speech and language difficulties from the age of 3, but some of them catch up and develop similarly to non-neglected children.

On Discrete Mixture of XLindley Distribution Using Lagrangian Probability Model: Properties and Applications in Count Data with Excess Zeros

On Discrete Mixture of XLindley Distribution Using Lagrangian Probability Model: Properties and Applications in Count Data with Excess Zeros

Trajectories of cardiovascular risk predict pregnancy outcomes: The Bogalusa Heart Study and the Cardiovascular Risk in Young Finns Study.

Life course patterns of change in risk-trajectories-affect health. To examine how trajectories of cardiovascular risk factors are associated with pregnancy and birth outcomes. Data from two cohort studies participating in the International Childhood Cardiovascular Consortium-The Bogalusa Heart Study (BHS; started in 1973, N = 903 for this analysis) and the Cardiovascular Risk in Young Finns Study (YFS; started in 1980, N = 499) were used. Both followed children into adulthood and measured cardiovascular risk factors, including body mass index (BMI), systolic and diastolic blood pressure (SBP/DBP), total, lipoprotein (LDL)- and high density lipoprotein (HDL)-cholesterol and serum triglycerides. Discrete mixture modelling was used to divide each cohort into distinct trajectories according to these risk factors from childhood to early adulthood, and these groups were then used to predict pregnancy outcomes including small for gestational age (SGA; <10th study-specific percentile of gestational age by sex), preterm birth (PTB; <37 weeks' gestation), hypertensive disorders of pregnancy (HDP) and gestational diabetes mellitus (GDM), with control for age at baseline and at first birth, parity, socioeconomic status, BMI and smoking. The models created more trajectories for BMI, SBP and HDL-cholesterol in the YFS than in BHS, for which three classes generally seemed to be sufficient to represent the groups in the population across risk factors. In BHS, the association between the higher and flatter DBP trajectory and PTB was aRR 1.77, 95% confidence interval [CI] 1.06, 2.96. In BHS the association between consistent total cholesterol and PTB was aRR 2.16, 95% CI 1.22, 3.85 and in YFS the association between elevated high trajectory and PTB was aRR 3.35, 95% CI 1.28, 8.79. Elevated-increasing SBP was associated with a higher risk of GH in BHS and increasing or persistent-obese BMI trajectories were associated with GDM in both cohorts (BHS: aRR 3.51, 95% CI 1.95, 6.30; YFS: aRR 2.61, 95% CI 0.96, 7.08). Trajectories of cardiovascular risk, particularly those that represent a consistent or more rapid worsening of cardiovascular health, are associated with a higher risk of pregnancy complications.

Online short text clustering using infinite extensions of discrete mixture models

AbstractShort text clustering is one of the fundamental tasks in natural language processing. Different from traditional documents, short texts are ambiguous and sparse due to their short form and the lack of recurrence in word usage from one text to another, making it very challenging to apply conventional machine learning algorithms directly. In this article, we propose two novel approaches for short texts clustering: collapsed Gibbs sampling infinite generalized Dirichlet multinomial mixture model infinite GSGDMM) and collapsed Gibbs sampling infinite Beta‐Liouville multinomial mixture model (infinite GSBLMM). We adopt two flexible and practical priors to the multinomial distribution where in the first one the generalized Dirichlet distribution is integrated, while the second one is based on the Beta‐Liouville distribution. We evaluate the proposed approaches on two famous benchmark datasets, namely, Google News and Tweet. The experimental results demonstrate the effectiveness of our models compared to basic approaches that use Dirichlet priors. We further propose to improve the performance of our methods with an online clustering procedure. We also evaluate the performance of our methods for the outlier detection task, in which we achieve accurate results.

Trajectories of self-reported pain-related health outcomes and longitudinal effects on medication use in rheumatoid arthritis: a prospective cohort analysis using the Australian Rheumatology Association Database (ARAD)

ObjectiveTo determine distinct trajectories of self-reported pain-related health status in rheumatoid arthritis (RA), their relationship with sociodemographic factors and medication use.Methods988 Australian Rheumatology Association Database participants with RA (71% female,...

Continuous Mixtures of Tractable Probabilistic Models

Probabilistic models based on continuous latent spaces, such as variational autoencoders, can be understood as uncountable mixture models where components depend continuously on the latent code. They have proven to be expressive tools for generative and probabilistic modelling, but are at odds with tractable probabilistic inference, that is, computing marginals and conditionals of the represented probability distribution. Meanwhile, tractable probabilistic models such as probabilistic circuits (PCs) can be understood as hierarchical discrete mixture models, and thus are capable of performing exact inference efficiently but often show subpar performance in comparison to continuous latent-space models. In this paper, we investigate a hybrid approach, namely continuous mixtures of tractable models with a small latent dimension. While these models are analytically intractable, they are well amenable to numerical integration schemes based on a finite set of integration points. With a large enough number of integration points the approximation becomes de-facto exact. Moreover, for a finite set of integration points, the integration method effectively compiles the continuous mixture into a standard PC. In experiments, we show that this simple scheme proves remarkably effective, as PCs learnt this way set new state of the art for tractable models on many standard density estimation benchmarks.

1094-P: Longitudinal Trajectories of Fasting C-Peptide in Youth-Onset Type 2 Diabetes—The SEARCH for Diabetes in Youth Study

In youth-onset type 2 diabetes (YO-T2D; diabetes diagnosed &amp;lt;20 years-old) loss of glycemic control is more rapid than in adults, partly attributed to accelerated beta cell decline. However, no studies have investigated the potential heterogeneity in the longitudinal trajectories of beta cell function and whether genetic factors might contribute to heterogeneity. We derived longitudinal trajectories of fasting C-peptide (FCP) levels, a surrogate measure of beta cell function, via discrete mixture modeling of repeated measures collected over 15 years of follow-up in youth with YO-T2D (antibody negative and insulin resistant; n=266) from the SEARCH for Diabetes in Youth cohort study. We tested the association between the derived trajectories and T2D genetic risk score (GRS) using a multinomial logistic regression model with adjustment for relevant covariates. A four-trajectory model best fit the data, classifying youth with YO-T2D as having either “initial low FCP and stable over time” (32%), “initial high FCP and slow incline” (8%), “initial high FCP and slow decline” (53%), or “initial very high FCP and rapid decline” (7%) (Figure). T2D GRS was not associated with FCP trajectory membership. These data demonstrate substantial heterogeneity in trajectories of FCP in YO-T2D and suggest that genetic predisposition to T2D is not a significant determinant of this heterogeneity. Disclosure A.Shapiro: None. A.Bellatorre: None. J.M.Stafford: None. A.S.Shah: None. C.Pihoker: None. F.Malik: None. A.D.Liese: None. D.J.Pettitt: None. D.Dabelea: None. Funding National Institute of Diabetes and Digestive and Kidney Diseases (1R01DK127208-01, 1UC4DK108173); Centers for Disease Control and Prevention (U18DP006131, U18DP006133, U18DP006134, U18DP006136, U18DP006138, U18DP006139)

0100 Developmental trajectories of SWA and ORP since childhood in young adults: association with psychiatric/behavioral disorders

Abstract Introduction Declines in non-rapid eye movement (NREM) sleep slow wave activity (SWA) in the transition to adolescence are indicative of brain maturational changes that parallel a decline in sleep depth, as measured by the novel sleep EEG biomarker odds ratio product (ORP). However, these developmental trajectories and their psychiatric/behavioral associations have lacked replication across multiple time points during the lifespan. Methods We determined the developmental trajectories (median follow-up 15y) of SWA and ORP among 150 participants from the Penn State Child Cohort studied in childhood (median age 9y) through adolescence (median age 16y) into young adulthood (median age 24y). Discrete mixture models for clustering of longitudinal data series was used to compute the probability of a subject belonging to mutually-exclusive 2 or 3 trajectory groups, which were selected based on Bayesian and Akaike Information Criteria. Cochran-Mantel-Haenszel tests were used to compare the proportion of subjects with psychiatric/behavioral disorders across the three developmental trajectories in each sleep EEG biomarker. Results Age-related declines in SWA were observed across two developmental groups, of which one followed a cubic trajectory and one a linear trajectory. This decline in SWA was coupled with age-related increases in ORP across three developmental groups, two of which were quadratic and one was linear. Compared to the shallowest trajectories, the SWA and ORP trajectories with the greatest slopes in the transition to young adulthood had a significantly higher proportion of participants with a history of psychiatric/behavioral disorders (78.6% vs. 53.2% and 80.0% vs. 50.7%, respectively, both p&amp;lt; 0.01). Conclusion Our novel data provide replication for the maturational trajectories of SWA and ORP by including time points across three distinct developmental stages, including young adulthood. Also, our data provide evidence linking the developmental trajectories of SWA and ORP with psychopathology. Young adults with psychiatric/behavioral disorders are associated with a steeper developmental decline in sleep intensity (SWA) and sleep depth (ORP) after adolescence. Support (if any) National Institutes of Health (R01HL136587, R01MH118308, UL1TR000127)

Virulence decomposition for bifurcating infections.

A fundamental goal in infection biology is to understand the emergence of variation in pathogen virulence-here defined as the decrease in host fitness caused by a pathogen. To uncover the sources of such variation, virulence can be decomposed into both host- and pathogen-associated components. However, decomposing virulence can be challenging owing to complex within-host pathogen dynamics such as bifurcating infections, which recently received increased empirical and theoretical attention. Bifurcating infections are characterized by the emergence of two distinct infection types: (i) terminal infections with high pathogen loads resulting in rapid host death, and (ii) persistent infections with lower loads and delayed host death. Here, we propose to use discrete mixture models to perform separate virulence decompositions for each infection type. Using this approach, we reanalysed a recently published experimental dataset on bacterial load and survival in Drosophila melanogaster. This analysis revealed several advantages of the new approach, most importantly the generation of a more comprehensive picture of the varying sources of virulence in different bacterial species. Beyond this application, our approach could provide valuable information for ground-truthing and improving theoretical models of within-host infection dynamics, which are developed to predict variation in infection outcome and pathogen virulence.

Discrete mixture representations of spherical distributions

We obtain discrete mixture representations for parametric families of probability distributions on Euclidean spheres, such as the von Mises–Fisher, the Watson and the angular Gaussian families. In addition to several special results we present a general approach to isotropic distribution families that is based on density expansions in terms of special surface harmonics. We discuss the connections to stochastic processes on spheres, in particular random walks, discrete mixture representations derived from spherical diffusions, and the use of Markov representations for the mixing base to obtain representations for families of spherical distributions.

Heterogeneity in choice experiment data: A Bayesian investigation

Discrete mixture (DM) models recognize the presence of heterogeneity across individuals in a given population. In the context of a public land use discrete choice experiment, we use DM models to allow for respondent behavior to probabilistically mix over multiple competing process heuristics. We pairwise combine the Random Utility Model (RUM), Contextual Concavity Model (CCM), and Random Regret Minimization (RRM) heuristic into three DM models, in which the probability of an individual adhering to a particular heuristic is modeled as a function of sociodemographic characteristics. We present a comprehensive Bayesian analysis for which we explicitly describe prior selection, inferential procedures, and model comparison metrics. We use a fully Bayesian information criterion to rank the models. We find evidence that responses are best modeled using random regret. After accounting for preference heterogeneity, the DM models estimate two latent groups of decision makers. For the DM models, we develop a novel algorithm to calculate posterior-weighted willingness to pay estimates for improvements in different public park amenities in Polk County, Iowa.

Physically Motivated Models for Heterogeneous Random Phenomena

In 2010, Tunde Ogunnaike published the book Random Phenomena, a comprehensive introduction to the physically based modeling of random phenomena for engineers. This paper is primarily a survey of concepts and methods for dealing with more complex heterogeneous random phenomena not addressed in Random Phenomena, but building on the foundation that the book establishes. Specifically, this paper discusses three extensions of that material: discrete mixture models, extensions of the linear regression models discussed in Random Phenomena (generalized linear models for strongly non-Gaussian response variables, and mixtures of regression models), and an extension of the Central Limit Theorem that is believed to be new. Since, as Random Phenomena emphasizes, software is necessary for analyzing and modeling random data, procedures in the open-source R programming environment are described that support all of the concepts and methods presented here.

Discrete Wavelet Transform-Based Gaussian Mixture Model for Remote Sensing Image Compression

High-ratio image compression is difficult because remote sensing images have complex background and rich information, and the correlation between features is weak. An accurate entropy model is an important way to solve the problem by enhancing the representation ability of the compression models. The entropy model is more suited to estimate the probability distributions with the sparse latent representations. This study proposes a novel entropy model (DWTGMM) based on discrete wavelet transform (DWT) and Gaussian mixture model (GMM) for remote sensing image compression. The method uses DWT to transform the latent representations into wavelet domain and obtains four sparse representations, and then uses the proposed DWTGMM to model them separately to estimate the probability distribution of each element. It is noteworthy that the DWT used in our approach does not require learning parameters and can be combined with other entropy models to acquire the distribution of latent representations. To evaluate our method, we construct three remote sensing image datasets, i.e., GoogleMap, GF1, and GF7. We compare our method with several popular learned compression models and traditional codecs. Experimental results show that the proposed method can achieve excellent performance with low complexity. Especially with the same model architecture, the DWTGMM achieves the best compression performance.