Skip to Main Content
 

Global Search Box

 
 
 

ETD Abstract Container

Abstract Header

Diversification and Generalization for Metric Learning with Applications in Neuroimaging

Shi, Bibo
http://orcid.org/0000-0002-9918-0448
http://rave.ohiolink.edu/etdc/view?acc_num=ohiou1448980736

Abstract Details

2015, Doctor of Philosophy (PhD), Ohio University, Computer Science (Engineering and Technology).
Many machine learning algorithms rely on “good” metrics to quantify the distances or similarities between data instances. Context dependent metrics learned from the training data are often e effective in improving the performance of metric-based algorithms under different circumstances, or for different tasks at hand. At present, most of existing metric learning algorithms learn metrics only from a binary similarity perspective, overlooking the fact that similarities tend to have different levels and binary configurations cannot fully account for many situations occurring in practice. In addition, many state-of-the-art metric learning solutions only estimate Mahalanobis metrics, which are linear transformation models with limited expressive power. More complicated nonlinear structures embedded in the data often cannot be well handled, thus failing to improve or even deteriorating the performance of the metric-based algorithm that follows. In this dissertation, we address the aforementioned drawbacks along two directions: diversify and generalize the forms of metric learning. For diversification, a novel Prior Distance Informed Metric Learning (PDIML) model is developed. Both global and local PDIML implementations have been successfully applied to diversify the similarities between data points through integrating prior distance knowledge into pairwise distance matrices. For generalization, we tackle metric learning from the perspective of feature transformation, and propose a set of novel nonlinear solutions through the utilization of deformable geometric models to learn spatially varying metrics. Thin-plate splines (TPS) are chosen as the geometric model due to their remarkable versatility and representation power in accounting for high-order deformations. TPS based metric learning algorithms have been developed for both k Nearest Neighbor (kNN) and Support Vector Machines (SVMs). Furthermore, a theoretically sound diffeomorphic constraint is proposed to ensure the legitimacy of metrics learned through our TPS models. Various experiments on both synthetic data sets and real world data sets, i.e., UCI repository, are conducted to validate the proposed models, with comparisons made against state-of-the-art solutions. Application-wise, we validate our PDIML model and TPS based models with 3D neuroimaging data, and utilize them to identify biomarkers for Alzheimer’s Disease (AD), and its early stage, Mild Cognitive Impairment (MCI). The data used in our analysis were obtained from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). A variety of image features, including patch-level gray matter concentration, longitudinal deformation magnitudes and structure-level volumes are extracted and selected. Experimental results demonstrate that our proposed metric learning solutions can effectively improve the classification accuracy for AD/MCI and normal control (NC).
Jundong Liu (Committee Chair)
Razvan Bunescu (Committee Member)
Cindy Marling (Committee Member)
Chang Liu (Committee Member)
Robert Colvin (Committee Member)
Martin Mohlenkamp (Committee Member)
126 p.
Computer Science
Machine Learning; Metric Learning; Neuroimaging; Thin-Plate Splines; ADNI

Recommended Citations

Citations

  • Shi, B. (2015). Diversification and Generalization for Metric Learning with Applications in Neuroimaging [Doctoral dissertation, Ohio University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=ohiou1448980736

    APA Style (7th edition)

  • Shi, Bibo. Diversification and Generalization for Metric Learning with Applications in Neuroimaging. 2015. Ohio University, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=ohiou1448980736.

    MLA Style (8th edition)

  • Shi, Bibo. "Diversification and Generalization for Metric Learning with Applications in Neuroimaging." Doctoral dissertation, Ohio University, 2015. http://rave.ohiolink.edu/etdc/view?acc_num=ohiou1448980736

    Chicago Manual of Style (17th edition)

ohiou1448980736
580
© 2015, all rights reserved.
This open access ETD is published by Ohio University and OhioLINK.