Penalized Preimage Learning in Kernel Principal Component Analysis

• Published in: IEEE transactions on neural networks Vol. 21; no. 4; pp. 551 - 570
• Main Authors: Wei-Shi Zheng, JianHuang Lai, Yuen, P.C.
• Format: Journal Article
• Language: English
• Published: New York, NY IEEE 01.04.2010; Institute of Electrical and Electronics Engineers
• Subjects: Algorithms; Applied sciences; Artificial Intelligence; Computer science; control theory; systems; Computer Simulation; Connectionism. Neural networks; Constraint optimization; Exact sciences and technology; Humans; Image denoising; Image Interpretation, Computer-Assisted - methods; Image reconstruction; Kernel; kernel principal component analysis (KPCA); Kernels; Laplace equations; Learning; Learning and adaptive systems; Lighting; locality preservation; Mathematical analysis; Mathematical models; Mean square error methods; Mutual information; Optimization; Pattern Recognition, Automated; Pattern recognition. Digital image processing. Computational geometry; penalty function; preimage problem; Principal Component Analysis; Turning; Vectors (mathematics); Occlusion; Image processing; Data mining; Modeling; Optimization; Image reconstruction; Mean square error; Penalty function; Facies; Illumination; preimage problem; Ill posed problem; kernel principal component analysis (KPCA); Convexity; Learning algorithm; Kernel; Locality; Image databank; Image restoration; Neural network; Laplacian; Kernel method; Image quality; Luminance; locality preservation; Artificial intelligence; Occultation; Mutual information; Facial expression; Principal component analysis

Finding the preimage of a feature vector in kernel principal component analysis (KPCA) is of crucial importance when KPCA is applied in some applications such as image preprocessing. Since the exact preimage of a feature vector in the kernel feature space, normally, does not exist in the input data...