当前位置:网站首页>论文笔记: 多标签学习 DM2L
论文笔记: 多标签学习 DM2L
2022-06-24 06:56:00 【闵帆】
摘要: 分享对论文的理解. 原文见 Ma, Z.-C., & Chen, S.-C. (2021). Expand globally, shrink locally: Discrimi-nant multi-label learning with missing labels. Pattern Recognition, 111, 107675.
1. 论文贡献
- 从全局和局部两个方面同时优化;
- 用核函数支撑非线性变换;
- 理论分析到位.
2. 基本符号
| 符号 | 含义 | 说明 |
|---|---|---|
| X ∈ R n × d \mathbf{X} \in \mathbb{R}^{n \times d} X∈Rn×d | 属性矩阵 | |
| X k ∈ R n k × d \mathbf{X}_k \in \mathbb{R}^{n_k \times d} Xk∈Rnk×d | 具有第 k k k 个标签的属性子矩阵 | |
| Y ∈ { − 1 , + 1 } n × c \mathbf{Y} \in \{-1, +1\}^{n \times c} Y∈{ −1,+1}n×c | 标签矩阵 | |
| Y ~ ∈ { − 1 , + 1 } n × l \tilde{\mathbf{Y}} \in \{-1, +1\}^{n \times l} Y~∈{ −1,+1}n×l | 观测到的标签矩阵 | |
| Ω = { 1 , … , n } × { 1 , … , c } \mathbf{\Omega} = \{1, \dots, n\} \times \{1, \dots, c\} Ω={ 1,…,n}×{ 1,…,c} | 观测标签位置集合 | |
| W ∈ R m × l \mathbf{W} \in \mathbb{R}^{m \times l} W∈Rm×l | 系数矩阵 | 仍然是线性模型 |
| w i ∈ R m \mathbf{w}_i \in \mathbb{R}^m wi∈Rm | 某一标签的系数向量 | |
| C ∈ R l × l \mathbf{C} \in \mathbb{R}^{l \times l} C∈Rl×l | 标签相关性矩阵 | 成对相关性, 不满足对称性 |
3. 算法
基本优化目标:
min 1 2 ∥ R Ω ( X W ) − Y ~ ∥ F 2 + λ d ∥ X W ∥ ∗ (1) \min \frac{1}{2} \|R_{\Omega}(\mathbf{XW}) - \tilde{\mathbf{Y}}\|_F^2 + \lambda_d \|\mathbf{XW}\|_*\tag{1} min21∥RΩ(XW)−Y~∥F2+λd∥XW∥∗(1)
其中,
- 损失函数部分不考虑缺失值, 这个属于常规操作.
- 核正则 (nuclear norm) 部分考虑了预测的矩阵, 而不仅仅是 X W \mathbf{XW} XW, 有点奇怪.
考虑标签结构后的优化目标:
min 1 2 ∥ R Ω ( X W ) − Y ~ ∥ F 2 + λ d ( ∑ k = 1 c ∥ X k W ∥ ∗ − ∥ X W ∥ ∗ ) , (2) \min \frac{1}{2} \|R_{\Omega}(\mathbf{XW}) - \tilde{\mathbf{Y}}\|_F^2 + \lambda_d \left(\sum_{k = 1}^c \|\mathbf{X}_k\mathbf{W}\|_* - \|\mathbf{XW}\|_*\right), \tag{2} min21∥RΩ(XW)−Y~∥F2+λd(k=1∑c∥XkW∥∗−∥XW∥∗),(2)
其中,
- ∥ X k W ∥ ∗ \|\mathbf{X}_k\mathbf{W}\|_* ∥XkW∥∗ 表达了局部标签结构, 轶越小越好;
- ∥ X W ∥ ∗ \|\mathbf{XW}\|_* ∥XW∥∗ 表达了全局标签结构, 轶越大越好 (可分性更强, 信息量越高).
- 这两点就是题目的来源.
增加非线性的优化目标:
min 1 2 ∥ R Ω ( X W ) − Y ~ ∥ F 2 + λ d ( ∑ k = 1 c ∥ f ( X k ) W ∥ ∗ − ∥ f ( X ) W ∥ ∗ ) , (5) \min \frac{1}{2} \|R_{\Omega}(\mathbf{XW}) - \tilde{\mathbf{Y}}\|_F^2 + \lambda_d \left(\sum_{k = 1}^c \|f(\mathbf{X}_k)\mathbf{W}\|_* - \|f(\mathbf{X})\mathbf{W}\|_*\right), \tag{5} min21∥RΩ(XW)−Y~∥F2+λd(k=1∑c∥f(Xk)W∥∗−∥f(X)W∥∗),(5)
其中 f ( ⋅ ) f(\cdot) f(⋅) 为核函数导致的非线性变换.
4. 小结
- 又是一堆理论证明.
边栏推荐
- Vscode topic recommendation
- Cold thinking on the hot track: multiplier effect is the fundamental requirement of East West calculation
- Search and recommend those things
- Examples of corpus data processing cases (reading multiple text files, reading multiple files specified under a folder, decoding errors, reading multiple subfolder text, batch renaming of multiple fil
- Teach you how to use the reflect package to parse the structure of go - step 1: parameter type check
- Ad-gcl:advantageous graph augmentation to improve graph contractual learning
- All you know is the test pyramid?
- VsCode主题推荐
- C language_ Love and hate between string and pointer
- Echart's experience (I): about y axis yaxis attribute
猜你喜欢

Swift foundation features unique to swift

The first exposure of Alibaba cloud's native security panorama behind the only highest level in the whole domain

Swift Extension ChainLayout(UI的链式布局)(源码)

Blue Bridge Cup_ Queen n problem

不止于观测|阿里云可观测套件正式发布

对于flex:1的详细解释,flex:1

Mousse shares listed on Shenzhen Stock Exchange: gross profit margin continued to decline, and marketing failed in the first quarter of 2022

Vulnhub target: boredhackerblog: social network

SCM stm32f103rb, BLDC DC motor controller design, schematic diagram, source code and circuit scheme

Shader common functions
随机推荐
Do you still have the opportunity to become a machine learning engineer without professional background?
How to cancel the display of the return button at the uniapp uni app H5 end the autobackbutton does not take effect
Screenshot recommendation - snipaste
Leetcode 207: course schedule (topological sorting determines whether the loop is formed)
C language_ Love and hate between string and pointer
[nilm] non intrusive load decomposition module nilmtk installation tutorial
GraphMAE----论文快速阅读
C语言_字符串与指针的爱恨情仇
基金的募集,交易与登记
redolog和binlog
C# Lambda
Atguigu---15- built in instruction
You get in Anaconda
Introduction to software engineering - Chapter 2 - feasibility study
JS implementation to check whether an array object contains values from another array object
工控机防破解
Latest news of awtk: new usage of grid control
[测试开发]初识软件测试
Industrial computer anti cracking
Swift Extension NetworkUtil(网络监听)(源码)