K-nearest neighbour
Introduction
1.
[Articals]
1.1.
[Hubs in Space: Popular Nearest Neighbors in High-Dimensional Data]
1.1.1.
[Introduction]
1.1.1.1.
The curse of dimensionality
1.1.1.2.
Notation
1.1.1.3.
What caused the skewness
1.1.2.
The Hubness Phenomenon
1.1.2.1.
A Motivating Example
1.1.2.2.
Hubness in Real Data
1.1.3.
The Origins of Hubness
1.1.3.1.
The Position of Hubs
1.1.3.2.
Mechanisms Behind Hubness
1.1.3.2.1.
why points closer to the data mean become hubs in high dimensions
1.1.3.3.
Hubness in Real Data
1.1.4.
The Impact of Hubness on Machine Learning
1.1.4.1.
Supervised Learning
1.1.4.1.1.
"GOOD" AND "BAD" k-OCCURRENCES
1.1.4.1.2.
"Bad" hubs
1.1.4.1.3.
Origin of "Bad" Hub
1.1.4.1.4.
INFLUENCE ON CLASSIFICATION ALGORITHMS
1.2.
自然最近邻居在高维数据结构学习中的应用
1.2.1.
摘要
1.2.2.
绪论
1.2.3.
自然最近邻居(3N)
1.2.3.1.
最近邻居技术
1.2.3.2.
自然最近邻居(3N)
1.2.3.3.
3N 邻居的特性
1.2.3.4.
构造自适应最近邻域图
1.2.3.5.
本章小结
1.2.4.
3N 邻居在谱图聚类中的应用
1.2.4.1.
聚类结构的子空间描述理论及方法
1.2.4.2.
改进的 MNCut 算法
1.2.4.3.
改进的 MNcut 算法：3N-MNcut 算法
1.3.
A Tutorial on Spectral Clustering
1.3.1.
Similarity graphs
1.3.2.
Graph Laplacians and their basic properties
1.3.3.
Spectral Clustering Algorithms
1.3.4.
Graph cut point of view
1.3.5.
Random walks point of view
1.3.6.
Practical details
1.3.7.
Outlook and further reading
1.4.
Nearest neighbor regression in the presence of bad hubs
1.5.
基于自然最近邻的无参聚类算法研究
1.5.1.
自然最近邻居的定义及改进
1.5.2.
基于自然最近邻居的无参聚类算法
2.
Visual Data Mining
2.1.
POIViz: a fast interactive method for visualizing a large collection of Open datasets
2.1.1.
Related Work
2.1.2.
CONTRIBUTION
2.2.
The Comparative Analysis of Major Foreign Visual Data Mining Software
2.2.1.
RAPIDMINER
2.3.
Improving Big Data Visual Analytics with Interactive Virtual Reality
3.
Seminar Log
4.
SpotLight
Powered by
GitBook
K-nearest neighbour
POIViz: a fast interactive method for visualizing a large collection of Open datasets