Title Participants Abstract
"Evaluating Testing, Profile Likelihood Confidence Interval Estimation, and Model Comparisons for Item Covariate Effects in Linear Logistic Test Models" "Sun-Joo Cho, Paul De Boeck, Woo-Yeol Lee" "© 2017, © The Author(s) 2017. The linear logistic test model (LLTM) has been widely applied to investigate the effects of item covariates on item difficulty. The LLTM was extended with random item residuals to account for item differences not explained by the item covariates. This extended LLTM is called the LLTM-R. In this article, statistical inference methods are investigated for these two models. Type I error rates and power are compared via Monte Carlo studies. Based on the simulation results, the use of the likelihood ratio test (LRT) is recommended over the paired-sample t test based on sum scores, the Wald z test, and information criteria, and the LRT is recommended over the profile likelihood confidence interval because of the simplicity of the LRT. In addition, it is concluded that the LLTM-R is the better general model approach. Inferences based on the LLTM while the LLTM-R is the true model appear to be largely biased in the liberal way, while inferences based on the LLTM-R while the LLTM is the true model are only biased in a very minor and conservative way. Furthermore, in the absence of residual variance, Type I error rate and power were acceptable except for power when the number of items is small (10 items) and also the number of persons is small (200 persons). In the presence of residual variance, however, the number of items needs to be large (80 items) to avoid an inflated Type I error and to reach a power level of.90 for a moderate effect."
"Probabilistic flood prediction for urban sub-catchments using sewer models combined with logistic regression models" "Xiaohan Li, Patrick Willems"
"Estimation of controlled direct effects on a dichotomous outcome using logistic structural direct effect models" "Stijn Vansteelandt" "We consider the problem of assessing whether an exposure affects a dichotomous outcome other than by modifying a given mediator. The standard approach, logistic regression adjusting for both exposure and the mediator, is known to be biased in the presence of confounders for the mediator-outcome relationship. Because additional regression adjustment for such confounders is only justified when they are not affected by the exposure, inverse probability weighting has been advocated, but is not ideally tailored to mediators that are continuous or have strong measured predictors. We overcome this limitation by developing inference for a novel class of causal models that are closely related to Robins' logistic structural direct effect models, but do not inherit their difficulties of estimation. We study identification and efficient estimation under the assumption that all confounders for the exposure-outcome and mediator-outcome relationships have been measured, and find adequate performance in simulation studies. We discuss extensions to case-control studies and relevant implications for the generic problem of adjustment for time-varying confounding."
"Comparison of spatial and aspatial logistic regression models for landmine risk mapping" "Craig Schultz, Aura Cecilia Alegria Caicedo, Jan Paul Herman Cornelis, Hichem Sahli" "Abstract Landmines continue to affect the lives of millions of people living in war-torn countries. One major challenge in humanitarian mine action (HMA) is finding new and integrated approaches to land release, which remains a slow and costly process. The use of geographic information systems (GIS) in HMA can improve the land release process by efficient mapping and prioritizing of landmine risk areas. This study explores the usage of aspatial and spatial regression techniques to construct a predictive geo-statistical model for landmine risk mapping in a small 160 km2 municipality in Bosnia and Herzegovina (BiH) and a large 4500 km2 region in Colombia. The first application of logistic geographically weighted regression to landmine risk mapping is presented. The results show that in the BiH study area, the effect of local parameters that influence the distribution of landmine risk varies significantly across the study area. Conversely, in the Colombia case study the effect of explanatory variables remains more homogeneous over the study area. We produced two landmine risk maps for each study area, based on aspatial and spatial regression models. Risk maps are classified into five classes, i.e. very low, low, medium, high, and very high risk. The landmine risk maps created through the usage of these innovative methodologies improve the assessment of risk and prioritization of the land release process in mine-contaminated areas, compared to existing approaches."
"Efficient and private scoring of decision trees, support vector machines and logistic regression models based on pre-computation" "Martine De Cock, Rafael Dowsley, Caleb Horst, Rajendra Katti, Anderson Nascimento, Wing-Sea Poon, Stacey Truex" "Many data-driven personalized services require that private data of users is scored against a trained machine learning model. In this paper we propose a novel protocol for privacy-preserving classification of decision trees, a popular machine learning model in these scenarios. Our solutions is composed out of building blocks, namely a secure comparison protocol, a protocol for obliviously selecting inputs, and a protocol for multiplication. By combining some of the building blocks for our decision tree classification protocol, we also improve previously proposed solutions for classification of support vector machines and logistic regression models. Our protocols are information theoretically secure and, unlike previously proposed solutions, do not require modular exponentiations. We show that our protocols for privacy-preserving classification lead to more efficient results from the point of view of computational and communication complexities. We present accuracy and runtime results for seven classification benchmark datasets from the UCI repository."
"Adnexal masses difficult to classify as benign or malignant using subjective assessment of gray scale and Doppler ultrasound findings: logistic regression models do not help" "Lieveke Ameye, Sabine Van Huffel, Dirk Timmerman" "AIM: To develop a logistic regression model that can discriminate between benign and malignant adnexal masses perceived to be difficult to classify by subjective evaluation of gray scale and Doppler ultrasound findings (subjective assessment) and to compare its diagnostic performance with that of subjective assessment, serum CA 125 and the risk of malignancy index (RMI). METHODS: We used the 3511 patients with an adnexal mass included in the International Ovarian Tumor Analysis (IOTA) studies. All patients had been examined with transvaginal gray scale and Doppler ultrasound following a standardized research protocol by an experienced ultrasound examiner using a high end ultrasound system. In addition to prospectively collecting information on > 40 clinical and ultrasound variables, the ultrasound examiner classified each mass as certainly or probably benign, unclassifiable, or certainly or probably malignant. A logistic regression model to discriminate between benignity and malignancy was developed for the unclassifiable masses (n = 244, i.e. 7% of all tumors) using a training set (160 tumors, 45 malignancies) and then tested on a test set (84 tumors, 28 malignancies). The gold standard was the histological diagnosis of the surgically removed adnexal mass. The area under the receiver operating characteristic curve (AUC), sensitivity, specificity, positive and negative likelihood ratio (LR+, LR-) were used to describe diagnostic performance and were compared between subjective assessment, CA 125, the RMI and the logistic regression model created. RESULTS: One variable was retained in the logistic regression model: the largest diameter (in mm) of the largest solid component of the tumor (OR 1.04, 95% CI 1.02 - 1.06). The model had an AUC of 0.68 (95% confidence interval, CI 0.59 to 0.78) on the training set and 0.65 (95%CI 0.53 to 0.78) on the test set. On the test set, a cutoff of 25% probability of malignancy (corresponding to largest diameter of largest solid component 23mm) resulted in sensitivity 64% (18/28), specificity 55% (31/56), LR+ 1.44 and LR- 0.65. The corresponding figures for subjective assessment were 68% (19/28), 59% (33/56), 1.65 and 0.55. On the test set of patients with available CA 125 results, the LR+ and LR- of the logistic regression model (cutoff 25% probability of malignancy) were 1.29 and 0.73, of subjective assessment 1.44 and 0.63, of CA 125 (cutoff 35 U/mL) 1.25 and 0.84 and of RMI (cutoff 200) 1.21 and 0.92. CONCLUSION: About 7% of adnexal masses that are considered appropriate to remove surgically cannot be classified as benign or malignant by experienced ultrasound examiners using subjective assessment. Logistic regression models to estimate the risk of malignancy, CA 125 measurements and the RMI are not helpful in these masses. Copyright © 2011 ISUOG. Published by John Wiley & Sons, Ltd."
"On the relationships between Jeffreys modal and weighted likelihood estimation of ability under logistic IRT models" "David Magis" "This paper focuses on two estimators of ability with logistic item response theory models: the Bayesian modal (BM) estimator and the weighted likelihood (WL) estimator. For the BM estimator, Jeffreys' prior distribution is considered, and the corresponding estimator is referred to as the Jeffreys modal (JM) estimator. It is established that under the three-parameter logistic model, the JM estimator returns larger estimates than the WL estimator. Several implications of this result are outlined. © 2011 The Psychometric Society."
"Developing risk models for multicenter data using standard logistic regression produced suboptimal predictions: A simulation study" "Ben Van Calster, Dirk Timmerman, Laure Wynants" "Although multicenter data are common, many prediction model studies ignore this during model development. The objective of this study is to evaluate the predictive performance of regression methods for developing clinical risk prediction models using multicenter data, and provide guidelines for practice. We compared the predictive performance of standard logistic regression, generalized estimating equations, random intercept logistic regression, and fixed effects logistic regression. First, we presented a case study on the diagnosis of ovarian cancer. Subsequently, a simulation study investigated the performance of the different models as a function of the amount of clustering, development sample size, distribution of center-specific intercepts, the presence of a center-predictor interaction, and the presence of a dependency between center effects and predictors. The results showed that when sample sizes were sufficiently large, conditional models yielded calibrated predictions, whereas marginal models yielded miscalibrated predictions. Small sample sizes led to overfitting and unreliable predictions. This miscalibration was worse with more heavily clustered data. Calibration of random intercept logistic regression was better than that of standard logistic regression even when center-specific intercepts were not normally distributed, a center-predictor interaction was present, center effects and predictors were dependent, or when themodel was applied in a new center. Therefore, to make reliable predictions in a specific center, we recommend random intercept logistic regression."
"Predictive performance of multinominal logistic prediction models" "Dirk Timmerman" "Multinomial Logistic Regression (MLR) has been advocated for developing clinical prediction models that distinguish between three or more unordered outcomes. We present a full-factorial simulation study to examine the predictive performance of MLR models in relation to the relative size of outcome categories, number of predictors and the number of events per variable. It is shown that MLR estimated by Maximum Likelihood yields overfitted prediction models in small to medium sized data. In most cases, the calibration and overall predictive performance of the multinomial prediction model is improved by using penalized MLR. Our simulation study also highlights the importance of events per variable in the multinomial context as well as the total sample size. As expected, our study demonstrates the need for optimism correction of the predictive performance measures when developing the multinomial logistic prediction model. We recommend the use of penalized MLR when prediction models are developed in small data sets or in medium sized data sets with a small total sample size (ie, when the sizes of the outcome categories are balanced). Finally, we present a case study in which we illustrate the development and validation of penalized and unpenalized multinomial prediction models for predicting malignancy of ovarian cancer."
"Validation and updating of risk models based on multinomial logistic regression." "Ben Van Calster" "BACKGROUND: Risk models often perform poorly at external validation in terms of discrimination or calibration. Updating methods are needed to improve performance of multinomial logistic regression models for risk prediction. METHODS: We consider simple and more refined updating approaches to extend previously proposed methods for dichotomous outcomes. These include model recalibration (adjustment of intercept and/or slope), revision (re-estimation of individual model coefficients), and extension (revision with additional markers). We suggest a closed testing procedure to assist in deciding on the updating complexity. These methods are demonstrated on a case study of women with pregnancies of unknown location (PUL). A previously developed risk model predicts the probability that a PUL is a failed, intra-uterine, or ectopic pregnancy. We validated and updated this model on more recent patients from the development setting (temporal updating; n = 1422) and on patients from a different hospital (geographical updating; n = 873). Internal validation of updated models was performed through bootstrap resampling. RESULTS: Contrary to dichotomous models, we noted that recalibration can also affect discrimination for multinomial risk models. If the number of outcome categories is higher than the number of variables, logistic recalibration is obsolete because straightforward model refitting does not require the estimation of more parameters. Although recalibration strongly improved performance in the case study, the closed testing procedure selected model revision. Further, revision of functional form of continuous predictors had a positive effect on discrimination, whereas penalized estimation of changes in model coefficients was beneficial for calibration. CONCLUSIONS: Methods for updating of multinomial risk models are now available to improve predictions in new settings. A closed testing procedure is helpful to decide whether revision is preferred over recalibration. Because multicategory outcomes increase the number of parameters to be estimated, we recommend full model revision only when the sample size for each outcome category is large."