Drexel University

About: Drexel University is a education organization based out in Philadelphia, Pennsylvania, United States. It is known for research contribution in the topics: Population & Poison control. The organization has 26770 authors who have published 51438 publications receiving 1949443 citations. The organization is also known as: Drexel & Drexel Institute.

Improving propensity score weighting using machine learning.

Abstract: Machine learning techniques such as classification and regression trees (CART) have been suggested as promising alternatives to logistic regression for the estimation of propensity scores. The authors examined the performance of various CART-based propensity score models using simulated data. Hypothetical studies of varying sample sizes (n=500, 1000, 2000) with a binary exposure, continuous outcome, and 10 covariates were simulated under seven scenarios differing by degree of non-linear and non-additive associations between covariates and the exposure. Propensity score weights were estimated using logistic regression (all main effects), CART, pruned CART, and the ensemble methods of bagged CART, random forests, and boosted CART. Performance metrics included covariate balance, standard error, per cent absolute bias, and 95 per cent confidence interval (CI) coverage. All methods displayed generally acceptable performance under conditions of either non-linearity or non-additivity alone. However, under conditions of both moderate non-additivity and moderate non-linearity, logistic regression had subpar performance, whereas ensemble methods provided substantially better bias reduction and more consistent 95 per cent CI coverage. The results suggest that ensemble methods, especially boosted CART, may be useful for propensity score weighting.

PassPoints: design and longitudinal evaluation of a graphical password system

Abstract: Computer security depends largely on passwords to authenticate human users. However, users have difficulty remembering passwords over time if they choose a secure password, i.e. a password that is long and random. Therefore, they tend to choose short and insecure passwords. Graphical passwords, which consist of clicking on images rather than typing alphanumeric strings, may help to overcome the problem of creating secure and memorable passwords. In this paper we describe PassPoints, a new and more secure graphical password system. We report an empirical study comparing the use of PassPoints to alphanumeric passwords. Participants created and practiced either an alphanumeric or graphical password. The participants subsequently carried out three longitudinal trials to input their password over the course of 6 weeks. The results show that the graphical password users created a valid password with fewer difficulties than the alphanumeric users. However, the graphical users took longer and made more invalid password inputs than the alphanumeric users while practicing their passwords. In the longitudinal trials the two groups performed similarly on memory of their password, but the graphical group took more time to input a password.

The Psychotherapy Dose‐Response Effect and Its Implications for Treatment Delivery Services

Abstract: To date, few studies have been published on the doseresponse relationship, but there is general consensus that between 13 and 18 sessions of therapy are required for 50% of patients to improve. Reviewing the clinical trials literature reveals that in carefully controlled and implemented treatments, between 57.6% and 67.2% of patients improve within an average of 12.7 sessions. Using naturalistic data, however, revealed that the average number of sessions received in a national database of over 6,000 patients was less than five. The rate of improvement in this sample was only about 20%. These results suggest that patients, on average, do not get adequate exposure to psychotherapy, nor do they recover from illness at rates observed in clinical trials research.

The crack problem for a nonhomogeneous plane

Abstract: The plane elasticity problem for a nonhomogeneous medium containing a crack is considered. It is assumed that the Poisson's ratio of the medium is constant and the Young's modulus E varies exponentially with the coordinate parallel to the crack. First the half plane problem is formulated and the solution is given for arbitrary tractions along the boundary. Then the integral equation for the crack problem is derived. It is shown that the integral equation having the derivative of the crack surface displacement as the density function has a simple Cauchy type kernel. Hence, its solution and the stresses around the crack tips have the conventional square root singularity. The solution is given for various loading conditions. The results show that the effect of the Poisson's ratio and consequently that of the thickness constraint on the stress intensity factors are rather negligible.

Formation of massive black holes through runaway collisions in dense young star clusters.

Abstract: A luminous X-ray source is associated with MGG 11—a cluster of young stars ∼200 pc from the centre of the starburst galaxy M 82 (refs 1, 2). The properties of this source are best explained3,4 by invoking a black hole with a mass of at least 350 solar masses (350 M⊙), which is intermediate between stellar-mass and supermassive black holes. A nearby but somewhat more massive cluster (MGG 9) shows no evidence of such an intermediate-mass black hole1,3, raising the issue of just what physical characteristics of the clusters can account for this difference. Here we report numerical simulations of the evolution and motion of stars within the clusters, where stars are allowed to merge with each other. We find that for MGG 11 dynamical friction leads to the massive stars sinking rapidly to the centre of the cluster, where they participate in a runaway collision. This produces a star of 800–3,000 M⊙, which ultimately collapses to a black hole of intermediate mass. No such runaway occurs in the cluster MGG 9, because the larger cluster radius leads to a mass segregation timescale a factor of five longer than for MGG 11.