Contributed Papers

Statistical Learning with Graphs and Networks

Wed, Aug 6, 8:30 AM - 10:20 AM Room CC-106B Music City Center

Keith LevinChair

Section on Statistical Learning and Data Science

7 Presentations

A Theoretical Framework to Signed Graph Learning

8:35 AM - 8:50 AM

Co-authors: Bisakh Banerjee (Michigan State University), Bisakh Banerjee (Michigan State University), Abdullah Karaaslanli (Department of Electrical and Computer Engineering, Michigan State University), Tapabrata Maiti (Michigan State University), Selin Aviyente (Department of Electrical and Computer Engineering, Michigan State University)

Graph Signal Processing Signed Graph Learning Estimation Error Graph Laplacian

An Ensemble Approach for Graph-Based Classification

8:50 AM - 9:05 AM

Co-authors: Jordan Eckert (Auburn University), Elvan Ceyhan (Auburn University), Jordan Eckert (Auburn University)

classification graph-based classifiers ensemble learning class imbalance boosting

Identifiability of Boolean graphical models

9:05 AM - 9:20 AM

Co-authors: Mengqi Lin, Gongjun Xu (University of Michigan), Mengqi Lin

Identifiability Boolean graphical models Graphical structures Boolean matrix decomposition Cognitive diagnosis models Hasse diagrams

Novel CCD-Based Algorithms for High-Accuracy Outlier Detection

9:20 AM - 9:35 AM

Co-authors: Rui Shi (Auburn University), Nedret Billor (Auburn University), Rui Shi (Auburn University), Elvan Ceyhan (Auburn University)

Outlier detection Outlyingness score Graph-based clustering Cluster catch digraphs High-dimensional data

Optimal Graph Joining and Graph Isomorphism

9:35 AM - 9:50 AM

Co-authors: Phuong Hoang (University of North Carolina at Charlotte), Yang Xiang (UNC Chapel Hill), Phuong Hoang (University of North Carolina at Charlotte), Bongsoo Yi (UNC Chapel Hill), Kevin McGoff, Andrew Nobel (University of North Carolina)

joinings optimal transport graph isomorphism random walks

Scalable Signed Exponential Random Graph Models under Local Dependence

9:50 AM - 10:05 AM

Co-authors: Marc Schalberger, Marc Schalberger, Cornelius Fritz (Trinity College Dublin)

Network Analysis Exponential Random Graph Models Signed Networks Local Dependence Large-Scale Networks

Statistical inference for noisy dynamic networks

10:05 AM - 10:20 AM

Co-authors: Peter MacDonald (University of Waterloo), Eric Kolaczyk (McGill University), Peter MacDonald (University of Waterloo)

Dynamic network Network inference Hidden Markov model Observation error