John R. Tipton, Mevin B. Hooten, Connor Nolan, Robert K. Booth, Jason McLachlan.

Source: The Annals of Applied Statistics, Volume 13, Number 4, 2363--2388.

Abstract:
Multivariate compositional count data arise in many applications including ecology, microbiology, genetics and paleoclimate. A frequent question in the analysis of multivariate compositional count data is what underlying values of a covariate(s) give rise to the observed composition. Learning the relationship between covariates and the compositional count allows for inverse prediction of unobserved covariates given compositional count observations. Gaussian processes provide a flexible framework for modeling functional responses with respect to a covariate without assuming a functional form. Many scientific disciplines use Gaussian process approximations to improve prediction and make inference on latent processes and parameters. When prediction is desired on unobserved covariates given realizations of the response variable, this is called inverse prediction. Because inverse prediction is often mathematically and computationally challenging, predicting unobserved covariates often requires fitting models that are different from the hypothesized generative model. We present a novel computational framework that allows for efficient inverse prediction using a Gaussian process approximation to generative models. Our framework enables scientific learning about how the latent processes co-vary with respect to covariates while simultaneously providing predictions of missing covariates. The proposed framework is capable of efficiently exploring the high dimensional, multi-modal latent spaces that arise in the inverse problem. To demonstrate flexibility, we apply our method in a generalized linear model framework to predict latent climate states given multivariate count data. Based on cross-validation, our model has predictive skill competitive with current methods while simultaneously providing formal, statistical inference on the underlying community dynamics of the biological system previously not available.

Emily Berg, Danhyang Lee.

Source: The Annals of Applied Statistics, Volume 13, Number 4, 2158--2188.

Abstract:
Quantiles of the distributions of several measures of erosion are important parameters in the Conservation Effects Assessment Project, a survey intended to quantify soil and nutrient loss on crop fields. Because sample sizes for domains of interest are too small to support reliable direct estimators, model based methods are needed. Quantile regression is appealing for CEAP because finding a single family of parametric models that adequately describes the distributions of all variables is difficult and small area quantiles are parameters of interest. We construct empirical Bayes predictors and bootstrap mean squared error estimators based on the linearly interpolated generalized Pareto distribution (LIGPD). We apply the procedures to predict county-level quantiles for four types of erosion in Wisconsin and validate the procedures through simulation.

Seong-Hwan Jun, Samuel W. K. Wong, James V. Zidek, Alexandre Bouchard-Côté.

Source: The Annals of Applied Statistics, Volume 13, Number 3, 1678--1707.

Abstract:
In this paper, we consider the knot-matching problem arising in computational forestry. The knot-matching problem is an important problem that needs to be solved to advance the state of the art in automatic strength prediction of lumber. We show that this problem can be formulated as a quadripartite matching problem and develop a sequential decision model that admits efficient parameter estimation along with a sequential Monte Carlo sampler on graph matching that can be utilized for rapid sampling of graph matching. We demonstrate the effectiveness of our methods on 30 manually annotated boards and present findings from various simulation studies to provide further evidence supporting the efficacy of our methods.

Irina Gaynanova.

Source: Bernoulli, Volume 26, Number 1, 286--322.

Abstract:
Sparse linear discriminant analysis via penalized optimal scoring is a successful tool for classification in high-dimensional settings. While the variable selection consistency of sparse optimal scoring has been established, the corresponding prediction and estimation consistency results have been lacking. We bridge this gap by providing probabilistic bounds on out-of-sample prediction error and estimation error of multi-class penalized optimal scoring allowing for diverging number of classes.

Daniela Pauger, Helga Wagner.

Source: Bayesian Analysis, Volume 14, Number 2, 341--369.

Abstract:
We propose a Bayesian approach to obtain a sparse representation of the effect of a categorical predictor in regression type models. As this effect is captured by a group of level effects, sparsity cannot only be achieved by excluding single irrelevant level effects or the whole group of effects associated to this predictor but also by fusing levels which have essentially the same effect on the response. To achieve this goal, we propose a prior which allows for almost perfect as well as almost zero dependence between level effects a priori. This prior can alternatively be obtained by specifying spike and slab prior distributions on all effect differences associated to this categorical predictor. We show how restricted fusion can be implemented and develop an efficient MCMC (Markov chain Monte Carlo) method for posterior computation. The performance of the proposed method is investigated on simulated data and we illustrate its application on real data from EU-SILC (European Union Statistics on Income and Living Conditions).

Alessandro Rinaldo, Ryan J. Tibshirani, Larry Wasserman.

Source: Statistical Science, Volume 34, Number 4, 599--603.

Abstract:
What is the meaning of a regression parameter? Why is this the de facto standard object of interest for statistical inference? These are delicate issues, especially when the model is misspecified. We argue that focusing on predictive quantities may be a desirable alternative.

Niko A. Busch
Jun 17, 2009; 29:7869-7876
BehavioralSystemsCognitive

PR Montague
Mar 1, 1996; 16:1936-1947
Articles

Gregor Thut
Sep 13, 2006; 26:9494-9502
BehavioralSystemsCognitive

It has been proposed that people can generate probabilistic predictions at multiple levels of representation during language comprehension. We used magnetoencephalography (MEG) and electroencephalography (EEG), in combination with representational similarity analysis, to seek neural evidence for the prediction of animacy features. In two studies, MEG and EEG activity was measured as human participants (both sexes) read three-sentence scenarios. Verbs in the final sentences constrained for either animate or inanimate semantic features of upcoming nouns, and the broader discourse context constrained for either a specific noun or for multiple nouns belonging to the same animacy category. We quantified the similarity between spatial patterns of brain activity following the verbs until just before the presentation of the nouns. The MEG and EEG datasets revealed converging evidence that the similarity between spatial patterns of neural activity following animate-constraining verbs was greater than following inanimate-constraining verbs. This effect could not be explained by lexical-semantic processing of the verbs themselves. We therefore suggest that it reflected the inherent difference in the semantic similarity structure of the predicted animate and inanimate nouns. Moreover, the effect was present regardless of whether a specific word could be predicted, providing strong evidence for the prediction of coarse-grained semantic features that goes beyond the prediction of individual words.

SIGNIFICANCE STATEMENT Language inputs unfold very quickly during real-time communication. By predicting ahead, we can give our brains a "head start," so that language comprehension is faster and more efficient. Although most contexts do not constrain strongly for a specific word, they do allow us to predict some upcoming information. For example, following the context of "they cautioned the...," we can predict that the next word will be animate rather than inanimate (we can caution a person, but not an object). Here, we used EEG and MEG techniques to show that the brain is able to use these contextual constraints to predict the animacy of upcoming words during sentence comprehension, and that these predictions are associated with specific spatial patterns of neural activity.

The action potential (AP) waveform controls the opening of voltage-gated calcium channels and contributes to the driving force for calcium ion flux that triggers neurotransmission at presynaptic nerve terminals. Although the frog neuromuscular junction (NMJ) has long been a model synapse for the study of neurotransmission, its presynaptic AP waveform has never been directly studied, and thus the AP waveform shape and propagation through this long presynaptic nerve terminal are unknown. Using a fast voltage-sensitive dye, we have imaged the AP waveform from the presynaptic terminal of male and female frog NMJs and shown that the AP is very brief in duration and actively propagated along the entire length of the terminal. Furthermore, based on measured AP waveforms at different regions along the length of the nerve terminal, we show that the terminal is divided into three distinct electrical regions: A beginning region immediately after the last node of Ranvier where the AP is broadest, a middle region with a relatively consistent AP duration, and an end region near the tip of nerve terminal branches where the AP is briefer. We hypothesize that these measured changes in the AP waveform along the length of the motor nerve terminal may explain the proximal-distal gradient in transmitter release previously reported at the frog NMJ.

SIGNIFICANCE STATEMENT The AP waveform plays an essential role in determining the behavior of neurotransmission at the presynaptic terminal. Although the frog NMJ is a model synapse for the study of synaptic transmission, there are many unknowns centered around the shape and propagation of its presynaptic AP waveform. Here, we demonstrate that the presynaptic terminal of the frog NMJ has a very brief AP waveform and that the motor nerve terminal contains three distinct electrical regions. We propose that the changes in the AP waveform as it propagates along the terminal can explain the proximal-distal gradient in transmitter release seen in electrophysiological studies.

Awareness for surprising sensory events is shaped by prior belief inferred from past experience. Here, we combined hierarchical Bayesian modeling with fMRI on an associative learning task in 28 male human participants to characterize the effect of the prior belief of tactile events on connections mediating the outcome of perceptual decisions. Activity in anterior insular cortex (AIC), premotor cortex (PMd), and inferior parietal lobule (IPL) were modulated by prior belief on unexpected targets compared with expected targets. On expected targets, prior belief decreased the connection strength from AIC to IPL, whereas it increased the connection strength from AIC to PMd when targets were unexpected. Individual differences in the modulatory strength of prior belief on insular projections correlated with the precision that increases the influence of prediction errors on belief updating. These results suggest complementary effects of prior belief on insular-frontoparietal projections mediating the precision of prediction during probabilistic tactile learning.

SIGNIFICANCE STATEMENT In a probabilistic environment, the prior belief of sensory events can be inferred from past experiences. How this prior belief modulates effective brain connectivity for updating expectations for future decision-making remains unexplored. Combining hierarchical Bayesian modeling with fMRI, we show that during tactile associative learning, prior expectations modulate connections originating in the anterior insula cortex and targeting salience-related and attention-related frontoparietal areas (i.e., parietal and premotor cortex). These connections seem to be involved in updating evidence based on the precision of ascending inputs to guide future decision-making.

While all eyes have been on the huge fall in the outright price of crude oil, the oil options markets...

The post Crude Oil Options Are Predicting Lasting Volatility appeared first on OpenMarkets.

Laurie Garrett, the prophet of this pandemic, expects years of death and “collective rage.”

The Bank of England says the UK faces its worst slump in 300 years, but on Thursday held off from any moves on rates or bond buying.

The National Children's Study, a large-scale, longitudinal, birth cohort study of US children that endeavors to identify preventable and environmental origins of chronic diseases, has begun recruitment.

In a highly diverse, urban setting, pregnant women can be recruited to participate in the National Children's Study at rates similar to those obtained in clinic settings. Refinements to the pregnancy screener and other components are needed to optimize implementation. (Read the full article)

Available clinical prediction rules to identify children with cerebrospinal fluid pleocytosis at low risk for Lyme meningitis include headache duration, cranial nerve palsy, and percent cerebrospinal fluid mononuclear cells. These rules require independent validation.

These clinical prediction rules accurately identify patients at low risk for Lyme meningitis in our large multicenter cohort. Children at low risk may be considered for outpatient management while awaiting Lyme serology. (Read the full article)

Previous research has demonstrated that young children with traumatic brain injury are at elevated risk of poor outcomes, particularly following severe injuries. These deficits persist until at least 5 years postinsult. Factors predicting outcomes in this age group have not been established.

This study follows survivors of very early traumatic brain injury into adolescence. Results indicate that severe injury is associated with poorest outcome, but after 3 years, the gap between children with severe traumatic brain injury and peers stabilizes. (Read the full article)

Survival of children born with Down syndrome has been improving, but few studies have used population-based data to examine the influence of fetal and maternal characteristics on survival.

This study examined predictors of survival for children born with Down syndrome using population-based data from the UK Northern Congenital Abnormality Survey and shows that year of birth, gestational age, birth weight, and presence of additional anomalies influence survival status. (Read the full article)

Early speech and language delays are risk factors for later developmental and social difficulties. It is easier to identify them retrospectively than prospectively. Population characteristics and prevalence rates make screening problematic.

Using data from a birth cohort, this study identifies predictors of language performance at 5 years and 4 patterns of change between 3 and 5 years, comparing those who change with those whose profile remains low across time points. (Read the full article)

Loud snoring, which spikes at ~2 to 3 years of age, has been associated with behavior problems in school-aged children in cross-sectional studies, but no longitudinal studies have quantified predictors and the behavioral impact of persistent snoring in preschool-aged children.

Persistent loud snoring, which occurs in 9% of children 2 to 3 years of age, is linked with behavior problems. Higher socioeconomic status and a history of breastfeeding were associated with lower rates of transient and persistent snoring in young children. (Read the full article)

Previous studies suggest that abnormal findings on conventional EEG during the neonatal period are associated with death or severe brain injury in preterm infants. However, large cohort studies on preterm EEG for predicting later neurodevelopmental outcome remain scarce.

This study demonstrates precise prognostic values of conventional EEG for predicting neurodevelopmental outcome in the current perinatal care setting. Additionally, its prognostic values are independent of severe injury on neuroimaging and clinical risk factors. (Read the full article)

Retinal hemorrhages occur in accidental and inflicted traumatic brain injury (ITBI) and some medical encephalopathies. Large numbers and peripherally located retinal hemorrhages are frequently cited as distinguishing features of ITBI in infants, but the predictive value has not been established.

This prospective retinal imaging study found that a diagnosis of ITBI in infants and children can be distinguished from other traumatic and nontraumatic causes by the presence of >25 dot-blot (intraretinal layer) hemorrhages (positive predictive value = 93%). (Read the full article)

Among middle-aged adults, genotype scores predict incident type 2 diabetes but do not improve prediction models based on clinical risk factors including family history and BMI. These clinical factors are more dynamic in adolescence, however.

A genotype score also predicts type 2 diabetes from adolescence over a mean 27 years of follow-up into adulthood but does not improve prediction models based on clinical risk factors assessed in adolescence. (Read the full article)

The past several decades have seen a dramatic increase in the costs of health care and the prevalence of childhood activity limitations. More families with children are experiencing financial burden related to the cost of health care and insurance.

We find significant inequities in the occurrence of delayed or forgone needed health care for families with children as a result of high health care–related financial burden and having a child with an activity limitation. (Read the full article)

Adolescents have high placebo response rates in depression treatment trials. Screening for depression will likely detect youth with a broad range of symptom severity, including some who would benefit from watchful waiting but might not require active treatment.

The strongest predictors of symptom persistence are depressive symptom severity at presentation and continued symptoms on repeat screening 6 weeks later. These results provide important information for the development of postscreening management protocols in the primary care setting. (Read the full article)

Autism is a disorder that significantly affects language/communication skills, with many children not developing fluent language. The rate of spoken language acquisition after severe language delay and predictors of functional language, beyond comorbid intellectual disability, is less clear.

This study uses the largest sample to date to examine the relationship between key deficits associated with autism and attainment of phrase and/or fluent speech after a severe language delay, providing information to guide therapeutic targets and developmental expectations. (Read the full article)

The American Academy of Pediatrics recommends neonates born at <37 weeks’ gestation receive a predischarge Infant Car Seat Challenge, meaning up to 500 000 infants qualify annually. However, little is known about incidence and risk factors for failure in this group.

This is the largest study to date to examine incidence and risk factors for failure of the Infant Car Seat Challenge. We sought to identify infants most at risk for failure to narrow the scope of testing. (Read the full article)

Several risk factors for both overweight and obesity in childhood are identifiable during infancy.

A simple risk algorithm can be used to quantify risk of overweight in children. It can be used to help identify at-risk infants in a clinical setting to facilitate targeted intervention. (Read the full article)

Extremely preterm infants are at high risk of neonatal mortality or morbidities. Existing prediction models focus on mortality, specific morbidities, or composite mortality and morbidity outcomes and ignore differences in outcome severity.

A simple and practical statistical model was developed that can be applied on the first day after NICU admission to predict outcome severity spanning from no morbidity to mortality. The model is highly discriminative (C-statistic = 90%) and internally valid. (Read the full article)

Approximately 50% of children diagnosed with attention-deficit/hyperactivity disorder (ADHD) at <7 years of age in the community do not meet criteria for ADHD over time. There is a need to examine predictors of diagnostic stability in young children with ADHD.

Predictors of diagnostic stability from early to middle childhood include child’s baseline externalizing and internalizing symptoms, parental history of psychopathology, and socioeconomic status. These predictors may guide treatment planning at the time of ADHD diagnosis. (Read the full article)

A large majority of preschool and young school age children with mental health problems do not receive services and little is known about the determinants of service use in this age group.

Behavioral, not emotional, disorders increase service use but only if impairment is present. Such impairment may operate via increased parental burden and parent and caregiver problem recognition. Low socioeconomic status has an independent effect increasing service use. (Read the full article)

Predicting high-risk populations in congenital diaphragmatic hernia (CDH) can help target care strategies. Prediction rules for infants with CDH often lack validation, are aimed at a prenatal population, and are of limited generalizability. We cannot currently discriminate the highest risk neonates during the crucial period shortly after birth.

This clinical prediction rule was developed and validated on an international database. It discriminates patients and high, intermediate, and low risk of mortality; is easy to apply; and is generalizable to most infants with CDH. (Read the full article)

Term MRI can assist in identifying the nature and extent of brain injury in preterm infants. However, brain injury detected by MRI does not fully account for neurodevelopmental impairments, particularly cognitive and behavioral impairments, common in preterm survivors.

In addition to brain injury, an assessment of brain growth by using one-dimensional measurements on MRI is helpful for predicting neurodevelopment. Two different patterns of impaired brain growth are observed that relate independently to early cognitive development in preterm infants. (Read the full article)

Lung ultrasound outperforms conventional radiology in the emergency diagnosis of pneumothorax and pleural effusions. In the pediatric age, lung ultrasound has been also successfully applied to the fluid-to-air transition after birth and to rapid pneumonia diagnosis.

Nasal ventilation has dramatically decreased the need for invasive mechanical respiratory support. This study demonstrates that, after a short trial on nasal continuous positive airway pressure, lung ultrasonography reliably predicts the failure of noninvasive ventilation unlike the conventional chest radiogram. (Read the full article)

Jaundiced newborns without additional risk factors rarely develop kernicterus if the total serum bilirubin is <25 mg/dL. Measuring the bilirubin/albumin ratio might improve risk assessment, but the relationships of both indicators to advancing stages of neurotoxicity are poorly documented.

Both total serum bilirubin and bilirubin/albumin ratio are strong predictors of advancing stages of acute and post-treatment auditory and neurologic impairment. However, bilirubin/albumin ratio, adjusted to the same sensitivity, does not improve prediction over total serum bilirubin alone. (Read the full article)

Pediatric Brain Injury Research Network investigators recently derived a highly sensitive clinical prediction rule for pediatric abusive head trauma (AHT).

The performance of this AHT screening tool has been validated. Four clinical variables, readily available at the time of admission, detect pediatric AHT with high sensitivity in intensive care settings. (Read the full article)

Fasting insulin levels in childhood are increasingly being used as a surrogate for insulin resistance and risk of later type 2 diabetes, despite only a moderate correlation with whole-body insulin sensitivity and few data related to adult outcomes.

Elevated insulin values between the ages of 3 and 6 years are associated with an elevated risk for later type 2 diabetes. In 9- to 18-year-olds, elevated BMI (but not insulin values) is associated with later type 2 diabetes. (Read the full article)

Obstructive sleep apnea syndrome (OSAS) is associated with significant comorbidity: behavioral problems, sleepiness, and impaired quality of life. However, the utility of OSAS symptoms versus polysomnography in the prediction of comorbidities or response to treatment is not well known.

Among children with OSAS, the Pediatric Sleep Questionnaire, a well-validated, simple 1-page symptom inventory, predicts key adenotonsillectomy-responsive OSAS comorbidities and their improvement after adenotonsillectomy. In contrast, polysomnographic results do not offer similar predictive value. (Read the full article)

Pediatric residents may not be achieving competency in neonatal intubation. Opportunities for intubation during residency are decreasing. A precise definition of competency during training is lacking.

Bayesian statistics may be used to describe neonatal intubation competency in residents. At least 4 successful intubations are needed to achieve competency. The first 2 intubation opportunities appear to predict how many intubation opportunities are ultimately needed to achieve competency. (Read the full article)

Discharge from the NICU requires coordination and may be delayed for nonmedical reasons. Predicting when patients will be medically ready for discharge can avoid these delays and result in cost savings for the hospital.

We developed a supervised machine learning approach using real-time patient data from the daily neonatology progress note to predict when patients will be medically ready for discharge. (Read the full article)

A previous multivariable statistical model, using individual patient data, estimated the probability of abusive head trauma based on the presence or absence of 6 clinical features: rib fracture, long-bone fracture, apnea, seizures, retinal hemorrhage, and head or neck bruising.

The model performed well in this validation, with a sensitivity of 72.3%, specificity of 85.7%, and area under the curve of 0.88. In children <3 years old with intracranial injury plus ≥3 features, the estimated probability of abuse is >81.5%. (Read the full article)

Guidelines for postdischarge monitoring of hyperbilirubinemia for neonates of white descent are available from the American Academy of Pediatrics; however, such information for healthy term and late preterm Chinese neonates is lacking.

A classification model for predicting the risk of significant hyperbilirubinemia in Chinese neonates was developed that combines a transcutaneous bilirubin–based nomogram with clinical risk factors. It classified newborns into 6 risk groups, which can guide clinicians in planning appropriate follow-up strategies. (Read the full article)

Objectives: Empiric antibiotic prescribing can be supported by guidelines and/or local antibiograms, but these have limitations. We sought to use data from a comprehensive electronic health record to use statistical learning to develop predictive models for individual antibiotics that incorporate patient-, and hospital-specific factors. This paper reports on the development and validation of these models on a large retrospective cohort.

Methods: This is a retrospective cohort study including hospitalized patients with positive urine cultures in the first 48 hours of hospitalization at a 1500 bed, tertiary care hospital over a 4.5 year period. All first urine cultures with susceptibilities were included. Statistical learning techniques, including penalized logistic regression, were used to create predictive models for cefazolin, ceftriaxone, ciprofloxacin, cefepime, and piperacillin-tazobactam. These were validated on a held-out cohort.

Results: The final dataset used for analysis included 6,366 patients. Final model covariates included demographics, comorbidity score, recent antibiotic use, recent antimicrobial resistance, and antibiotic allergies. Models had acceptable to good discrimination in the training dataset and acceptable performance in the validation dataset, with a point estimate for area under the receiver operating characteristic curve (AUC) that ranged from 0.65 for ceftriaxone to 0.69 for cefazolin. All models had excellent calibration.

Conclusion: In this study we used electronic health record data to create predictive models to estimate antibiotic susceptibilities for UTIs in hospitalized patients. Our models had acceptable performance in a held-out validation cohort.

BACKGROUND:

Congenital diaphragmatic hernia (CDH) is a rare congenital anomaly with a mortality of ~27%. The Congenital Diaphragmatic Hernia Study Group (CDHSG) developed a simple postnatal clinical prediction rule to predict mortality in newborns with CDH. Our aim for this study is to externally validate the CDHSG rule in the European population and to improve its prediction of mortality by adding prenatal variables.

BACKGROUND AND OBJECTIVES:

Children born very preterm (VPT) are at high risk of cognitive impairment that impacts their educational and social opportunities. This study examined the predictive accuracy of assessments at 2, 4, 6, and 9 years in identifying preterm children with cognitive impairment by 12 years.

Maureen Hack
Aug 1, 2005; 116:333-341
ARTICLES

Andrew T. Shennan
Oct 1, 1988; 82:527-532
ARTICLES

Vinod K. Bhutani
Jan 1, 1999; 103:6-14
ARTICLES