Machine learning based on sparsity-inducing regularization for matrices

基于稀疏诱导矩阵正则化的机器学习

• 批准号: 22700138
• 负责人: TOMIOKA Ryota
• 金额: $ 2.58万
• 依托单位: The University of Tokyo
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Young Scientists (B)
• 财政年份: 2010
• 资助国家: 日本
• 起止时间: 2010 至 2012
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/grant/KAKENHI-PROJECT-22700138/
• 关键词: 機械学習; スパース性; 正則化; 行列; テンソル; スパース正則化; スパース; 最適化

The outcomes of this research project can be summarized as follows:1. I have extended the dual augmented Lagrangian (DAL) algorithm to deal with spectral regularization for matrices and proposed the M-DAL algorithm(ICML2010). The super-linear convergence of DAL and M-DAL algorithms wasproven and published in JMLR. A review of DAL and related algorithms has beenpublished as part of “Optimization for Machine Learning” (MIT Press). I have alsomade the code publicly available to promote its use in wider research communities. 2. In order to deal with non-numerical data, I have extended DAL to handle multiplekernel learning with thousands of kernels. This was published in MachineLearning Journal. 3. I have extended the framework to spectral regularization for higher-order tensors and analyzed its statistical performance. This was presented at NIPS2011.

本研究的主要成果如下：1.本文将对偶增广拉格朗日（DAL）算法推广到矩阵谱正则化，提出了M-DAL算法（ICML 2010）。DAL和M-DAL算法的超线性收敛性已在JMLR上得到证明并发表。DAL和相关算法的综述已经作为“机器学习优化”（MIT出版社）的一部分发表。我还公开了代码，以促进其在更广泛的研究社区中的使用。2.为了处理非数值数据，我扩展了DAL来处理具有数千个内核的多核学习。这篇文章发表在MachineLearning Journal上。3.我已经扩展的框架，高阶张量的谱正则化，并分析其统计性能。这是在NIPS 2011上提出的。

项目成果

A Bayesian Analysis of the Radioactive Releases of Fukushima.

福岛放射性释放的贝叶斯分析。

• 发表时间: 2012
• 作者: 田川和義;長谷川恭子;備藤達郎;小森優;来見良誠;森川茂廣;田中覚;李周浩;平井慎一;田中弘美;R. Tomioka and M. Morup
• 通讯作者: R. Tomioka and M. Morup

Discovering Emerging Topics in Social Streams via Link-Anomaly Detection

• DOI: 10.1109/tkde.2012.239
• 发表时间: 2014-01-01
• 期刊: IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
• 影响因子: 8.9
• 作者: Takahashi, Toshimitsu;Tomioka, Ryota;Yamanishi, Kenji
• 通讯作者: Yamanishi, Kenji

A Fast Augmented Lagrangian Algorithm for Learning Low-Rank Matrices

• 发表时间: 2010-06
• 期刊: Applied Physics Letters
• 影响因子: 4
• 作者: Ryota Tomioka;Taiji Suzuki;Masashi Sugiyama;H. Kashima

Augmented Lagrangian Methods for Learning, Selecting, and Combining Features

• DOI: 10.7551/mitpress/8996.003.0011
• 发表时间: 2011
• 作者: Ryota Tomioka;Taiji Suzuki;Masashi Sugiyama

SpicyMKL: a fast algorithm for Multiple Kernel Learning with thousands of kernels

• DOI: 10.1007/s10994-011-5252-9
• 发表时间: 2011-06
• 期刊: Machine Learning
• 影响因子: 7.5
• 作者: Taiji Suzuki;Ryota Tomioka