Applications Of Statistical Learning To Image Super-resolution And Tree-shaped Data

Statistical Learning is a discipline that based on data to establish probability and statistical model and conducting forecast and analysis.Data science is developing rapidly today,statistical learning has lots of applications in all walks of life.The main content of this paper could be divided into three parts:1.A Boosting Method to Face Image Super-resolution.The conventional sparsity-based methods enforce sparse coding on face image patches and such a sparse coding model regularizes all facial patches equally,which however ignores distinct natures of different facial patches for image reconstruction.In this paper,we propose a new weighted-patch super-resolution method based on AdaBoost,called W P A A algorithm(Weighted-Patch Algorithm via AdaBoost).Various experimental results on standard face database show that WPAA outperforms state-of-the-art methods in terms of both objective metrics and visual quality.2.Discriminative Super-resolution Gaussian Processes.The discriminative informa-tion about face image is of great importance.In this paper,the discriminative information is learned via Gaussian Processes and introduced into face super-resolution to propose a new approach,DSRGP(Discriminative Super-Resolution via Gaussian Processes)3.The Correlation Between Two Tree-Shaped Data.In this paper,the statistical model is established for tree-shaped data,and two methods,point estimation and MCMC algorithm in Bayesian statistics,are used to conduct statistical analysis of correlation between two tree-shaped data.