Big Data Lecture Series

Thursday, October 10, 2019

"Efficient Estimation of Functional Linear Models under Local Box Cox Transformation"

Dr. Mohammed Chowdhury, Assistant Professor of Statistics at Kennesaw State University
12:30pm - 1:45pm, Clendenin Building, room 1009, Kennesaw Campus (PDF)

ABSTRACT: An efficient method to estimate functional linear model has been proposed and developed for the longitudinal data under the setting of Local Box-Cox transformation (LBCT). When one Box-Cox transformation is not enough to approximate big longitudinal data, multiple Box-Cox transformations on response variable by some time (age) variable should be used for approximation to normality. These multiple Box-Cox transformations will be called as Local Box Transformation (BCT). To apply LBCT, we adopt a three-step estimation procedure. First, we split the longitudinal data by some time (age) variable. We then apply LBCT on each data in second step, and in third step, we accomplish efficient estimation of our models by incorporating three nonparametric smoothers, known as local polynomial smoother, kernel smoother and spline smoother. An application of our method has been demonstrated by using NGHS (National Growth and Health Study) longitudinal data.

BIO: Mohammed Chowdhury is an Assistant Professor in the Department of Statistics and Analytical Sciences at Kennesaw State University since Fall 2015. He earned his Ph.D. in Statistics from The George Washington University in 2013. After his Ph.D., he worked 9 months at K12 as an Analytic Researcher. He did one year Post-Doctoral research on Biomedical Sciences at National Institute of Health (NIH) before joining at Kennesaw State University. His research interest are on Nonparametric Regression, Bayesian Analysis, Time Series Analysis and Biostatistics.

  • Thursday, November 21, 2019

    "Interpretable and Integrative Deep Learning for Survival Analysis Using Histopathological Images and Genomic Data"

    Jie Hao, Postdoctoral researcher in the Perelman School of Medicine at the University of Pennsylvania; Graduated from KSU's Ph.D. in Analytics and Data Science (July 2019)
    12:30pm - 1:30pm, Clendenin Building, room 1009 on the Kennesaw Campus

    ABSTRACT: The integration of multi-modal data, such as histopathological images and genomic data, is essential for understanding cancer heterogeneity and complexity for personalized treatments, as well as for enhancing survival predictions in cancer study. Histopathology, as a clinical gold-standard tool for diagnosis and prognosis in cancers, allows clinicians to make precise decisions on therapies, whereas high-throughput genomic data have been investigated to dissect the genetic mechanisms of cancers. We propose a biologically interpretable deep learning model (PAGE-Net) that integrates histopathological images and genomic data, not only to improve survival prediction, but also to identify genetic and histopathological patterns that cause different survival rates in patients. PAGE-Net consists of pathology/genome/demography-specific layers, each of which provides comprehensive biological interpretation. In particular, we propose a novel patch-wise texture-based convolutional neural network, with a patch aggregation strategy, to extract global survival-discriminative features, without manual annotation for the pathology-specific layers. We adapted the pathway-based sparse deep neural network, named Cox-PASNet, for the genome-specific layers. The proposed deep learning model was assessed with the histopathological images and the gene expression data of Glioblastoma Multiforme (GBM) at The Cancer Genome Atlas (TCGA) and The Cancer Imaging Archive (TCIA). PAGE-Net achieved a C-index of 0.702, which is higher than the results achieved with only histopathological images (0.509) and Cox-PASNet (0.640). More importantly, PAGE-Net can simultaneously identify histopathological and genomic prognostic factors associated with patients’ survivals. The source code of PAGE-Net is publicly available at

    Photo of Jie HaoBIO: Jie Hao is a postdoctoral researcher in the Perelman School of Medicine at the University of Pennsylvania since August 2019. Jie just earned the PhD in Analytics and Data Science at the Kennesaw State University in July 2019. Her research interests are developing innovative computational methodologies in biomarker discovery, survival analysis, interpretable deep learning modeling, data integration of heterogeneous data, and handling missing data.