没有合适的资源?快使用搜索试试~ 我知道了~
本文档详细阐释了将深度学习相关理论成果应用于SVM使SVM机器学习效果性能更加优良
资源推荐
资源详情
资源评论
Machine Learning
I
Machine learning algorithms are very useful for many
applications (Classification, Regression, Adaptive Control)
I
They use datasets or experiences to fit a model
I
Example applications are:
I
Object Recognition
I
Face Recognition
I
Handwritten Digit Recognition
I
fMRI-Scan Classification
I
Medical Diagnosis
I
Document Classification
I
Many Applications in Bio-informatics and Chemistry
Marco A. Wiering 3/25
Limitations of Support Vector Machines
I
Support Vector Machines (SVM) often outperform other
machine learning methods
I
However, the standard SVM has a single adjustable layer
of weights
I
Instead of using such “shallow models”, deep architectures
can be better alternatives
I
SVMs use a-priori chosen kernel functions to compute
similarities between input vectors
I
A problem is that the choice of kernel function is important,
but kernel functions are not very flexible
I
Therefore we propose the deep SVM (DSVM)
I
The DSVM contains multiple layers of SVMs
Marco A. Wiering 4/25
Support Vector Regression
I
The objective function of the SVM is based on structural
risk minimization theory developed by Vapnik in the 1960s
I
Goal: find g(x) most suitable to the data, e.g. for
regression, -insensitive (Hinge) loss function:
I
|y
i
− g(x
i
)| ≤ ε
I
But also generalize well!
I
g(x) as flat as possible ⇒ ||w|| as small as possible
I
Yields a convex optimization problem
Marco A. Wiering 5/25
剩余24页未读,继续阅读
资源评论
finishspan
- 粉丝: 1
- 资源: 1
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
最新资源
- 筷手引流工具.apk
- 论文(最终)_20240430235101.pdf
- 基于python编写的Keras深度学习框架开发,利用卷积神经网络CNN,快速识别图片并进行分类
- 最全空间计量实证方法(空间杜宾模型和检验以及结果解释文档).txt
- 5uonly.apk
- 蓝桥杯Python组的历年真题
- 2023-04-06-项目笔记 - 第一百十九阶段 - 4.4.2.117全局变量的作用域-117 -2024.04.30
- 2023-04-06-项目笔记 - 第一百十九阶段 - 4.4.2.117全局变量的作用域-117 -2024.04.30
- 前端开发技术实验报告:内含4四实验&实验报告
- Highlight Plus v20.0.1
资源上传下载、课程学习等过程中有任何疑问或建议,欢迎提出宝贵意见哦~我们会及时处理!
点击此处反馈
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功