Tests For Constancy Of Model Parameters Over Time

2002 ◽  
Vol 14 (1-2) ◽  
pp. 113-132 ◽  
Author(s):  
Nils Lid Hjort ◽  
Alexander Koning
Keyword(s):  
Model Parameters ◽  
Over Time

scholarly journals Dynamic Detection of Delayed Cerebral Ischemia Using Machine Learning

2020 ◽  
Author(s):  
Murad Megjhani ◽  
Kalijah Terilli ◽  
Ayham Alkhachroum ◽  
David J. Roh ◽  
Sachin Agarwal ◽  
...  
Keyword(s):  
Machine Learning ◽  
Cerebral Ischemia ◽  
Characteristic Curve ◽  
Delayed Cerebral Ischemia ◽  
Risk Scores ◽  
Support Vector ◽  
Model Parameters ◽  
Learning Approaches ◽  
Physiologic Data ◽  
Over Time

AbstractObjectiveTo develop a machine learning based tool, using routine vital signs, to assess delayed cerebral ischemia (DCI) risk over time.MethodsIn this retrospective analysis, physiologic data for 540 consecutive acute subarachnoid hemorrhage patients were collected and annotated as part of a prospective observational cohort study between May 2006 and December 2014. Patients were excluded if (i) no physiologic data was available, (ii) they expired prior to the DCI onset window (< post bleed day 3) or (iii) early angiographic vasospasm was detected on admitting angiogram. DCI was prospectively labeled by consensus of treating physicians. Occurrence of DCI was classified using various machine learning approaches including logistic regression, random forest, support vector machine (linear and kernel), and an ensemble classifier, trained on vitals and subject characteristic features. Hourly risk scores were generated as the posterior probability at time t. We performed five-fold nested cross validation to tune the model parameters and to report the accuracy. All classifiers were evaluated for good discrimination using the area under the receiver operating characteristic curve (AU-ROC) and confusion matrices.ResultsOf 310 patients included in our final analysis, 101 (32.6%) patients developed DCI. We achieved maximal classification of 0.81 [0.75-0.82] AU-ROC. We also predicted 74.7 % of all DCI events 12 hours before typical clinical detection with a ratio of 3 true alerts for every 2 false alerts.ConclusionA data-driven machine learning based detection tool offered hourly assessments of DCI risk and incorporated new physiologic information over time.

scholarly journals Learning Personalized Itemset Mapping for Cross-Domain Recommendation

2020 ◽  
Author(s):  
Yinan Zhang ◽  
Yong Liu ◽  
Peng Han ◽  
Chunyan Miao ◽  
Lizhen Cui ◽  
...  
Keyword(s):  
Single Domain ◽  
Model Parameters ◽  
The Real ◽  
Cross Domain ◽  
Proposed Model ◽  
Latent Space ◽  
Learning Parameter ◽  
Generation Procedure ◽  
Over Time

Cross-domain recommendation methods usually transfer knowledge across different domains implicitly, by sharing model parameters or learning parameter mappings in the latent space. Differing from previous studies, this paper focuses on learning explicit mapping between a user's behaviors (i.e. interaction itemsets) in different domains during the same temporal period. In this paper, we propose a novel deep cross-domain recommendation model, called Cycle Generation Networks (CGN). Specifically, CGN employs two generators to construct the dual-direction personalized itemset mapping between a user's behaviors in two different domains over time. The generators are learned by optimizing the distance between the generated itemset and the real interacted itemset, as well as the cycle-consistent loss defined based on the dual-direction generation procedure. We have performed extensive experiments on real datasets to demonstrate the effectiveness of the proposed model, comparing with existing single-domain and cross-domain recommendation methods.

scholarly journals Bayesian B-spline mapping for dynamic quantitative traits

2012 ◽  
Vol 94 (2) ◽  
pp. 85-95 ◽  
Author(s):  
JUN XING ◽  
JIAHAN LI ◽  
RUNQING YANG ◽  
XIAOJING ZHOU ◽  
SHIZHONG XU
Keyword(s):  
Legendre Polynomial ◽  
Legendre Polynomials ◽  
Simulated Dataset ◽  
Model Parameters ◽  
B Spline ◽  
B Splines ◽  
Bayesian Mapping ◽  
Bayesian Shrinkage ◽  
Dynamic Traits ◽  
Over Time

SummaryOwing to their ability and flexibility to describe individual gene expression at different time points, random regression (RR) analyses have become a popular procedure for the genetic analysis of dynamic traits whose phenotypes are collected over time. Specifically, when modelling the dynamic patterns of gene expressions in the RR framework, B-splines have been proved successful as an alternative to orthogonal polynomials. In the so-called Bayesian B-spline quantitative trait locus (QTL) mapping, B-splines are used to characterize the patterns of QTL effects and individual-specific time-dependent environmental errors over time, and the Bayesian shrinkage estimation method is employed to estimate model parameters. Extensive simulations demonstrate that (1) in terms of statistical power, Bayesian B-spline mapping outperforms the interval mapping based on the maximum likelihood; (2) for the simulated dataset with complicated growth curve simulated by B-splines, Legendre polynomial-based Bayesian mapping is not capable of identifying the designed QTLs accurately, even when higher-order Legendre polynomials are considered and (3) for the simulated dataset using Legendre polynomials, the Bayesian B-spline mapping can find the same QTLs as those identified by Legendre polynomial analysis. All simulation results support the necessity and flexibility of B-spline in Bayesian mapping of dynamic traits. The proposed method is also applied to a real dataset, where QTLs controlling the growth trajectory of stem diameters in Populus are located.

scholarly journals A Bayesian Approach to Estimating Reciprocal Effects with the Bivariate STARTS Model using Markov Chain Monte Carlo

2021 ◽  
Author(s):  
Oliver Lüdtke ◽  
Alexander Robitzsch ◽  
Esther Ulitzsch
Keyword(s):  
Bayesian Approach ◽  
Prior Distribution ◽  
Parameter Estimates ◽  
Model Parameters ◽  
Reciprocal Effects ◽  
Ml Estimation ◽  
Model Parameterization ◽  
Lagged Effects ◽  
The Bayesian Approach ◽  
Over Time

The bivariate Stable Trait, AutoRegressive Trait, and State (STARTS) model provides a general approach for estimating reciprocal effects between constructs over time. However, previous research has shown that this model is difficult to estimate using the maximum likelihood (ML) method (e.g., nonconvergence). In this article, we introduce a Bayesian approach for estimating the bivariate STARTS model and implement it in the software Stan. We discuss issues of model parameterization and show how appropriate prior distributions for model parameters can be selected. Specifically, we propose the four-parameter beta distribution as a flexible prior distribution for the autoregressive and cross-lagged effects. Using a simulation study, we show that the proposed Bayesian approach provides more accurate estimates than ML estimation in challenging data constellations. An example is presented to illustrate how the Bayesian approach can be used to stabilize the parameter estimates of the bivariate STARTS model.

scholarly journals Memory patterns for choice adaptation in dynamic environments

2020 ◽  
Author(s):  
Emmanouil Konstantinidis ◽  
Jason L. Harman ◽  
Cleotilde Gonzalez
Keyword(s):  
Computational Models ◽  
Future Research ◽  
Model Parameters ◽  
Data Sets ◽  
Individual Memory ◽  
Learning And Adaptation ◽  
Decision Environments ◽  
Optimal Adaptation ◽  
Over Time

An important aspect of making good decisions is the ability to adapt to changes in the values of available options. Research suggests that we are poor at changing behavior and adapting our choices successfully. This work contributes to clarifying the role of memory on learning and successful adaptation to changing decision environments. We test the effects of the direction of change and the type of feedback using a decisions from experience binary choice task, where individuals learn the outcomes and their associated probabilities from feedback received after selecting between available choice options. The results revealed a robust effect of the direction of change: risk that becomes more rewarding over time is harder to detect than risk that becomes less rewarding over time; and even with full information about the outcomes of choice options people showed sub-optimal adaptation to change. We rely on three distinct computational models to interpret the role of memory on learning and adaptation. The distributions of individual model parameters were analyzed in relation to participants' ability to successfully adapt to the changing conditions of the various decision environments. Consistent across the three models and two distinct data sets (our experimental data and other researchers' data), results revealed the value of recency as an individual memory component for choice adaptation. Individuals relying more on recent experiences were more successful at adapting to change, regardless of the direction of change. We explain the value and limitations of these findings as well as opportunities for future research.

scholarly journals RESIDUAL ANALYSIS IN RASCH POISSON COUNTS MODELS

2021 ◽  
Vol 39 (1) ◽  
pp. 206
Author(s):  
Naiara Caroline Aparecido dos SANTOS ◽  
Jorge Luiz BAZÁN
Keyword(s):  
Item Response Theory ◽  
Item Response ◽  
Bayesian Approach ◽  
Residual Analysis ◽  
Model Parameters ◽  
Irt Model ◽  
Latent Traits ◽  
Over Time ◽  
Dichotomous Item ◽  
Poisson Counts

A Rasch Poisson counts (RPC) model is described to identify individual latent traits and facilities of the items of tests that model the error (or success) count in several tasks over time, instead of modeling the correct responses to items in a test as in the dichotomous item response theory (IRT) model. These types of tests can be more informative than traditional tests. To estimate the model parameters, we consider a Bayesian approach using the integrated nested Laplace approximation (INLA). We develop residual analysis to assess model t by introducing randomized quantile residuals for items. The data used to illustrate the method comes from 228 people who took a selective attention test. The test has 20 blocks (items), with a time limit of 15 seconds for each block. The results of the residual analysis of the RPC were promising and indicated that the studied attention data are not well tted by the RPC model.

scholarly journals Hybrid systems approach to modeling stochastic dynamics of cell size

2016 ◽  
Author(s):  
Cesar Augusto Vargas-Garcia ◽  
Abhyudai Singh
Keyword(s):  
Growth Rate ◽  
Cell Size ◽  
Size Variation ◽  
Model Parameters ◽  
Constant Growth Rate ◽  
Division Rate ◽  
Size Dependent ◽  
Daughter Cells ◽  
Constant Growth ◽  
Over Time

A ubiquitous feature of all living cells is their growth over time followed by division into two daughter cells. How a population of genetically identical cells maintains size homeostasis, i.e., a narrow distribution of cell size, is an intriguing fundamental problem. We model size using a stochastic hybrid system, where a cell grows exponentially over time and probabilistic division events are triggered at discrete time intervals. Moreover, whenever these events occur, size is randomly partitioned among daughter cells. We first consider a scenario, where a timer (i.e., cell-cycle clock) that measures the time since the last division event regulates cellular growth and the rate of cell division. Analysis reveals that such a timer-driven system cannot achieve size homeostasis, in the sense that, the cell-to-cell size variation grows unboundedly with time. To explore biologically meaningful mechanisms for controlling size we consider three different classes of models: i) a size-dependent growth rate and timer-dependent division rate; ii) a constant growth rate and size-dependent division rate and iii) a constant growth rate and division rate that depends both on the cell size and timer. We show that each of these strategies can potentially achieve bounded intercellular size variation, and derive closed-form expressions for this variation in terms of underlying model parameters. Finally, we discuss how different organisms have adopted the above strategies for maintaining cell size homeostasis.

Adjusting a subject-specific time of event in longitudinal studies

2019 ◽  
Vol 29 (7) ◽  
pp. 1787-1798
Author(s):  
Hyunkeun Ryan Cho ◽  
Seonjin Kim ◽  
Myung Hee Lee
Keyword(s):  
Goodness Of Fit ◽  
Real Life ◽  
Model Parameters ◽  
Specific Time ◽  
Goodness Of Fit Test ◽  
Proposed Model ◽  
Subject Specific ◽  
The Subject ◽  
The Mean ◽  
Over Time

Biomedical studies often involve an event that occurs to individuals at different times and has a significant influence on individual trajectories of response variables over time. We propose a statistical model to capture the mean trajectory alteration caused by not only the occurrence of the event but also the subject-specific time of the event. The proposed model provides a post-event mean trajectory smoothly connected with the pre-event mean trajectory by allowing the model parameters associated with the post-event mean trajectory to vary over time of the event. A goodness-of-fit test is considered to investigate how well the proposed model is fit to the data. Hypothesis tests are also developed to assess the influence of the subject-specific time of event on the mean trajectory. Theoretical and simulation studies confirm that the proposed tests choose the correctly specified model consistently and examine the effect of the subject-specific time of event successfully. The proposed model and tests are also illustrated by the analysis of two real-life data from a biomarker study for HIV patients along with their own time of treatment initiation and a body fatness study in girls with different age of menarche.

On the existence of a lactate threshold during incremental exercise: a systems analysis

1996 ◽  
Vol 80 (5) ◽  
pp. 1819-1828 ◽  
Author(s):  
M. E. Cabrera ◽  
H. J. Chizeck
Keyword(s):  
Experimental Data ◽  
Threshold Model ◽  
Lactate Concentration ◽  
Incremental Exercise ◽  
Recursive Least Squares ◽  
System Response ◽  
Model Parameters ◽  
Phase System ◽  
Time Varying ◽  
Over Time

The relationship between blood lactate concentration ([La]) and O2 uptake (VO2) during incremental exercise remains controversial: does [La] increase smoothly as a function of VO2 (continuous model), or does it begin to increase abruptly above a particular metabolic rate (threshold model)? The dynamic characteristics of the underlying physiological system are investigated using system identification analysis techniques. A multivariate deterministic time series model of the [La] and VO2 response to incremental changes in work rate was fitted to simulated and experimental data. Time-varying system response parameters were determined through the application of a weighted recursive least squares algorithm. The model, using the identified time-varying parameters, provided a good fit to the data. The variation of these parameters over time was then examined. Two major transitions in the parameters were found to occur at intensity levels equivalent to 53 +/- 8% and 77 +/- 9% maximal VO2 (experimental data). These changes in the model parameters indicate that the best linear dynamic model that fits the observed system behavior has changed. This implies that the system has changed its operation in some way, by altering its structure or by moving to a different operating region. The identified parameter changes over time suggest that the exercise intensity range (from rest to maximal VO2) is divided into three main intensity domains, each with distinct dynamics. Further study of this three-phase system may help in the understanding of the underlying physiological mechanisms that affect the dynamics of [La] and VO2 during exercise.

Visceral organ growth patterns in Saanen goats

2019 ◽  
Vol 157 (9-10) ◽  
pp. 721-742 ◽  
Author(s):  
M. E. B. Andrade ◽  
C. J. Härter ◽  
M. Gindri ◽  
K. T. Resende ◽  
I. A. M. A. Teixeira
Keyword(s):  
Body Weight ◽  
Large Intestine ◽  
Growth Curve ◽  
Inflection Point ◽  
Growth Patterns ◽  
Model Parameters ◽  
Visceral Organ ◽  
Intact Male ◽  
Organ Growth ◽  
Over Time

AbstractVisceral organs play an important role in animals' energy requirements, so their growth must be well understood. The objective of the current study was to fit and compare growth curves that best describe body and visceral organ growth over time in Saanen goats of different sexes. Data were synthesized from seven studies in which curves were fitted to visceral organ growth over time for female, intact male and castrated male Saanen goats from 5 to 45 kg body weight. The liver, pancreas, spleen, rumen–reticulum, omasum, abomasum, small intestine, large intestine and mesenteric adipose tissue (MAT) data were fitted to eight models: simple linear regression, quadratic, monomolecular, Brody, Von Bertalanffy, logistic, Gompertz and Richards. The best-fit model was chosen based on the corrected Akaike information criterion and the concordance correlation coefficient. Model parameters for each sex were compared. Overall, the model that best described visceral organ growth was the logistic model. Sex did not influence the parameters that predicted organ growth (g), except for MAT, where females presented a lower tissue deposition rate and greater inflection point than males. Irrespective of sex, at the beginning of the growth curve, the liver accounted for 28 ± 1.1 g/kg of empty body weight, and the inflection point occurred at 1.7 months. The rumen–reticulum and large intestine presented higher growth rates in the first 2 months of life. Knowledge of the visceral organ growth curve is useful in improving the understanding of the effect of nutritional requirements for goats and must be used to optimize the nutritional plans.

Sign in / Sign up

Export Citation Format

Share Document