請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/19147
標題: | L1正規化的平滑近似方法之比較 A Comparison of Smoothing Approximation Methods for L1-regularized LR |
作者: | Cheng-Xia Chang 常成霞 |
指導教授: | 林智仁(Chih-Jen Lin) |
關鍵字: | L1正則化,平滑近似, L1-approximation,Unconstrained Optimization, |
出版年 : | 2016 |
學位: | 碩士 |
摘要: | L1正規化的分類器被廣泛應用于獲取稀疏模型,但是其二階不可導特性對優化過程帶來了很大的挑戰。本文中,消除非約束性優化方法的使用限制,通過平滑近似L1的方法使其二階可微,從而可以使用常見的牛頓法解決。進一步探討該方法的應用空間,我們將之與L1和L2正規化問題的最佳化方法做了詳細的對比實驗,結果證實平滑化近似法繼承了L1和L2的某些特性,但應用前景依然不容樂觀。 L1-regularized classi ers are widely used to obtain sparse models; however, thenon-di erentiability of L1-regularized form causes more challenges in optimization. In this thesis, in order to eliminate the limited use of standard unconstrained methods in L1-regularized problems, smooth convex pproximations are used to replace the absolute form to make the problem twice-di erentiable. Thus we can use popular Newton methods to solve the reformulated problem. Results show that after approximation, the modified problem behaves in the middle of L1- and L2-regularized problems. To further investigate the pratical application of this approximation method, we also conduct experiments to compare with the state-of-art methods for olving L1- and L2-regularization problems. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/19147 |
DOI: | 10.6342/NTU201602127 |
全文授權: | 未授權 |
顯示於系所單位: | 資訊網路與多媒體研究所 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-105-1.pdf 目前未授權公開取用 | 2.53 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。