Key words: Deep Neural Network, Hierarchical Classification, Label Noise
Chen, L., Huang, N., Mu, C., Helm, H. S., Lytvynets, K., Yang, W., & Priebe, C. E. (2022). Deep Learning with Label Noise: A Hierarchical Approach. [Preprint]
Priebe, C. E., Huang, N., Villar, S., Mu, C., & Chen, L. (2022). Deep Learning is Provably Robust to Symmetric Label Noise. [Preprint]
Dynamic Network Sampling for Community Detection
Design dynamic network sampling scheme to optimize block recovery for stochastic blockmodel in the case where it is expensive to observe the entire graph; provide justification for using Chernoff information in subsequent inference for graphs.
Key words: Dynamic Network Sampling, Stochastic Blockmodel, Community Detection, Chernoff Information
Mu, C., Park, Y., & Priebe, C. E. (2023). Dynamic Network Sampling for Community Detection. Applied Network Science. [Paper][Preprint]
Community Detection for Stochastic Blockmodel Graphs with Vertex Covariates
Develop model-based spectral algorithms for clustering vertices in stochastic blockmodel graphs with vertex covariates; assess effects of observed and unobserved vertex heterogeneity on block recovery; employ Chernoff information to analytically compare the algorithms' performance and derive the Chernoff ratio expression for certain models of interest.
Key words: Spectral Graph Inference, Community Detection, Stochastic Blockmodel, Vertex Covariates, Chernoff Ratio
Mu, C., Mele, A., Hao, L., Cape, J., Athreya, A., & Priebe, C. E. (2022). On Spectral Algorithms for Community Detection in Stochastic Blockmodel Graphs with Vertex Covariates. IEEE Transactions on Network Science and Engineering. [Paper][Preprint]
Statistical Models for Large Networks
Built statistical models and algorithms that could be scaled to analyze large networks; estimated and simulated network formation models using high performance computing; developed R package with research purposes such as identifying the community structure.
Key words: (Generalized) Random Dot Product Graph, Stochastic Blockmodel with Covariates, Parallel Computation
Developed automatic tools for analyzing and annotating video stream with relevant Information such as timing, speed, traffic, accidents, objects and etc.
Key words: Structural Similarity Index, Earth Mover’s Distance, Oriented FAST and Rotated BRIEF, Image Hashing, Robust Image Similarity Measure, Deep Neural Networks
Mu, C., & Budavári, T. (2018). Dash Cam Video Analysis: Laptimes and Beyond. Poster presented at 2018 IDIES Annual Symposium, Baltimore, MD. [Poster]
Therapy Functional Measures
Identified patterns in patient functional trajectories; measured causal effect of different physical therapy dosage regimes on patient functional status; constructed features and built model to predict AMPAC score to optimize physical therapy in the hospital.
Key words: Linear Mixed-Effect Model, ARIMA
Crockett M., Mu, C., & Dahbura, A. T. (2018). Predictive Analytics for Patient Mobility Using AM-PAC. Poster presented at 2018 Johns Hopkins Research Symposium on Engineering in Healthcare, Baltimore, MD. [Poster]
Text Mining and Information Extraction
Collaborated with different teams to mine the large-scale text data, speculated gender based on names and explored characteristic distribution across gender; extracted information from large-scale data sets and reconstructed data; crawled online data.
Key words: Parallel Computation, Regular Expression, Crawler, Data Wrangling