Dai, Juan. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2007. / Includes bibliographical references (leaves 36-39). / Abstracts in English and Chinese. / Abstract --- p.i / Acknowledgement --- p.iii / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Previous Work --- p.3 / Chapter 1.2 --- Proposed Framework --- p.5 / Chapter 1.3 --- Overview --- p.6 / Chapter 2 --- Placement of the Local Classifiers --- p.8 / Chapter 2.1 --- The Uncertainty Map --- p.9 / Chapter 2.2 --- Responsibility Mixture Model --- p.11 / Chapter 2.3 --- EM for Parameter Estimation --- p.12 / Chapter 2.3.1 --- E-Step --- p.14 / Chapter 2.3.2 --- M-Step --- p.15 / Chapter 2.3.3 --- Relationship with Gaussian Mixture Mod- els(GMM) --- p.16 / Chapter 3 --- Fusing of Locally Adaptive Classifiers --- p.18 / Chapter 3.1 --- Training --- p.18 / Chapter 3.2 --- Testing --- p.21 / Chapter 4 --- Algorithmic Characteristics --- p.23 / Chapter 4.1 --- Uncertainty Piloted Placement of Local Classifiers --- p.23 / Chapter 4.2 --- Uncertainty Piloted Fusing of Local Classifiers --- p.24 / Chapter 4.3 --- Related Work --- p.25 / Chapter 5 --- Experiments --- p.27 / Chapter 5.1 --- Dimensionality Reduction --- p.27 / Chapter 5.2 --- Two-Class Classification Problem: Gender Classification --- p.29 / Chapter 5.3 --- Multi-Class Classification: Face Recognition --- p.30 / Chapter 5.3.1 --- Varying the Lighting --- p.31 / Chapter 5.3.2 --- Varying the Pose --- p.32 / Chapter 5.3.3 --- Number of Features Extracted --- p.33 / Chapter 6 --- Conclusion --- p.34 / Bibliography --- p.36

Photocopy of typescript. / Digitized by Kansas Correctional Industries