Meetings of the Eastern North American Region of the International Biometric Society (a.k.a. "ENAR meetings") are held in late March or early April each year and reflect the broad interests of the Society, including both quantitative techniques and application areas. Faculty and student presenters from the Department of Biostatistics regularly participate giving invited talks, contributed talks, and poster presentations.
The Joint Statistical Meetings, known simply as "JSM", is the largest gathering of statisticians held annually in North American. Faculty and student presenters from the Department of Biostatistics regularly participate giving invited talks, contributed talks, and poster presentations. Our students often receive top awards and participate in the affiliated career marketplace at the event.
Biostatistics guest speaker, Wensheng Guo, University of Pennsylvania, will present, "Dynamic functional clustering with applications to Multidisciplinary Approach to the Study of Chronic Pelvic Pain (MAPP) study."
In this talk, I will present some works under development by our group on functional clustering and dimension reduction methods. The proposed methods are motivated by, and are applied to the data collected in the NIDDK-funded Multidisciplinary Approach to the Study of Chronic Pelvic Pain (MAPP) Research Network. This is a longitudinal cohort study that collects high dimensional longitudinal urological chronic pelvic pain syndrome (UCPPS) symptom data together with many other biomarkers, neuroimaging data and genetic data. The goal of the study is to identify risk factors that can predict whether a subject would be worsening or improving and to understand the underlying pathological mechanisms. We first propose a dynamic functional clustering algorithm where each group of curves are modeled by a functional mixed effects model, and the posterior probability is used to iteratively classify each subject into different subgroups. The functional mixed effects model allows flexible designs and nested structures. The classification takes into account both group-average trajectories and between-subject variability. We propose an equivalent dynamic state space model to calculate the likelihood in fitting the model, and to efficiently compute the posterior probability in classifying a new subject. The resultant sequential algorithm is O(n) and can be implemented online. We also propose a leave-one-subject-out cross-validation Kullback-Leibler information criterion to choose the number of clusters. The performance is assessed through a simulation study, and we apply the proposed methods to longitudinal urological chronic pelvic pain syndrome symptom data collected in the MAPP Research Network and identify three subgroups
Last Updated On Wednesday, March 28, 2018 by Temp, GSPH Marketing & Development
Created On Friday, January 5, 2018
Click to enter calendar events or share school news and announcements.