487 research outputs found
Nonparametric Bayesian Dictionary Learning for Analysis of Noisy and Incomplete Images
No full text1 online resource (PDF, 29 pages, includes illustrations)Zhou, Mingyuan; Chen, Haojun; Paisley, John; Ren, Lu; Li, Lingbo; Xing, Zhengming; Dunson, David; Sapiro, Guillermo; Carin, Lawrence. (2010). Nonparametric Bayesian Dictionary Learning for Analysis of Noisy and Incomplete Images. Retrieved from the University Digital Conservancy, https://hdl.handle.net/11299/180858
Editorial: Ionizing radiation reprograms tumor immune microenvironment by inducing immunogenic cell death.
No full textInference of Low-Dimensional Latent Structure in High-Dimensional Data
No full textThe problem of learning a latent model for sparse or low-dimensional representation of high-dimensional data has attracted significant attention for many years. This thesis focuses on learning latent models for sparse or low-dimensional representation of images, dynamic data, and documents with Bayesian nonparametrics. The thesis consists of three parts.First, nonparametric Bayesian methods are considered for recovery of imagery based upon compressive measurements. A truncated beta-Bernoulli process is employed to infer an appropriate dictionary for the test data, and also for image recovery. In the context of compressive sensing, significant improvements in image recovery are manifested using learned dictionaries, relative to using standard orthonormal image expansions. The compressive-measurement projections are also optimized for the learned dictionary. Spatial inter-relationships within imagery are exploited through use of the Dirichlet and probit stick-breaking processes. Several example results are presented, with comparisons to other state-of-the-art methods in the literature.Second, hierarchical Bayesian methods are employed to learn a reversible statistical embedding. The proposed embedding procedure is connected to spectral embedding methods, for example, diffusion maps and Isomap, yielding a new statistical spectral framework. The proposed approach allows one to discard the training data when embedding new data, allows synthesis of high-dimensional data from the embedding space, and provides accurate estimation of the latent-space dimensionality. Hierarchical Bayesian methods are also developed to learn a nonlinear dynamic model in the low-dimensional embedding space, allowing joint analysis of multiple types of dynamic data, sharing strength and inferring inter-relationships. In addition to analyzing dynamic data, the learned model also yields effective synthesis. Example results are presented for statistical embedding, latent-space dimensionality estimation, and analysis and synthesis of high-dimensional (dynamic) motion-capture data.Third, a new hierarchical tree-based topic model is developed, based on nonparametric Bayesian techniques. The model has two unique attributes: (i) a child node in the tree may have more than one parent, with the goal of eliminating redundant sub-topics deep in the tree; and (ii) parsimonious sub-topics are manifested, by removing redundant usage of words at multiple scales. The depth and width of the tree are unbounded within the prior, with a retrospective sampler employed to adaptively infer the appropriate tree size based upon the corpus under study. Excellent quantitative results are manifested on five standard data sets, and the inferred tree structure is also found to be highly interpretable.</p
Human Driver Simulation Model
No full textAs part of the vehicle environmental certification process, the Environmental Protection Agency (EPA) requires automobile manufacturers to run a series of “drive cycle” tests to evaluate the efficiency of a vehicle (miles/gallon for internal combustion engine vehicles or Wh/mile for electric vehicles). For these tests, the dynamometer must be controlled by a human driver. The goal of this project is to create a simulation model of a human driver performing an automobile speed control task using MATLAB and Simulink. This model mimics human control tendencies and error as closely as possible by tuning parameter values to best-fit experimental data. Outputs from the model are brake pedal and accelerator pedal positions. Inputs to the model are the current desired vehicle speed, the current actual vehicle speed, and the desired vehicle speed a short time in the future. This preview of the desired speed is available to human drivers in the dynamometer test, and including preview as a control pathway in the simulation model was critical to producing reasonable results.
Keywords – Simulation, Manual control, Parameter optimization, Grey-box mode
Associations of Traumatic Injury with Abnormal Glucose Metabolism: A Population-Based Prospective Cohort Study
No full textTao Liu,1,* Xin Liu,1,* Yue Li,1 Aitian Wang,2 Shuohua Chen,3 Shouling Wu,3 Shike Hou,1 Haojun Fan,1 Chunxia Cao1 1Institute of Disaster and Emergency Medicine, Tianjin University, Tianjin, People’s Republic of China; 2Department of Intensive Medicine, Kailuan General Hospital, Tangshan, People’s Republic of China; 3Department of Cardiology, Kailuan General Hospital, Tangshan, People’s Republic of China*These authors contributed equally to this workCorrespondence: Chunxia Cao; Haojun Fan, Institute of Disaster and Emergency Medicine, Tianjin University, No. 92 Weijin Road, Nankai District, Tianjin, 300072, People’s Republic of China, Tel +86 02227893596, Fax +86 02227893596-307, Email [email protected]; [email protected]: Empirical data on the association between traumatic injury and abnormal glucose metabolism risk is limited. This study aimed to investigate the association between traumatic injury and abnormal glucose metabolism.Patients and Methods: This study included 153,162 participants in the Kailuan Study from 2006 to 2013. Participants with abnormal glucose metabolism at baseline were excluded. All participants were monitored every two years until December 31, 2019. During follow-up, 1915 subjects with a first traumatic injury (defined as a physical injury caused by an external force) were identified. For each subject with traumatic injury, one control subject was randomly selected and matched for age (± 3 years) and sex. A total of 3830 subjects were included in the final analysis. Cox proportional hazards models were used to examine the association between traumatic injury and the subsequent risk of abnormal glucose metabolism.Results: During a median follow-up of 6.91 (3.57– 9.41) years, 990 abnormal glucose metabolism events occurred. After adjustment for demographics, lifestyle behaviors, and traditional risk factors, those who had traumatic injury compared to controls were 32% more likely to develop any abnormal glucose metabolism (hazard ratio [HR] 1.32; 95% confidence interval [CI]1.16– 1.49), including impaired fasting glucose (IFG) (HR 1.29; 95% CI 1.12– 1.48) and diabetes (HR 1.37; 95% CI 1.10– 1.70). The risks for abnormal glucose metabolism, IFG, and diabetes in subjects with moderate-severe injury were higher than in subjects with mild injury for the 1-year follow-up period, while the association was not significantly different by injury severity for the whole follow-up period.Conclusion: Traumatic injury was associated with an increased risk of abnormal glucose metabolism. However, the risks of outcome events decreased as the follow-up period extended. Improved short- and long-term prevention and management strategies for controlling glucose are needed for individuals with traumatic injury.Keywords: traumatic injury, glucose metabolism, diabetes, impaired fasting glucose, cohort stud
- …
