Intro

This is a course project of Data Mining (2024, fall) in THU.

Objective

Implementing clustering on the human activity dataset.

Data

6 kinds of activities with 561 features https://archive.ics.uci.edu/dataset/240/human+activity+recognition+using+smartphones

Content

1. data analysis and preprocessing

2. clustering analysis

2.1 algorithms

2.1.1 Partitioning Methods

K-means
CLARANS

2.1.2 Hierarchical Clustering Methods

Agglomerative Hierarchical Clustering
Divisive Hierarchical Clustering

2.1.3 Density-Based Clustering Methods

DBSCAN

2.1.4 Grid-Based Clustering Methods

2.2 evaluation

2.2.1 clustering tendency

Hopkins statistic

2.2.2 clustering quality

Internal metrics
- Compatness
- Separation
- lhouette score
- Calinski-Harabasz index (CH index)
- avies-Bouldin index (DB index)
External metrics
- Adjusted Rand Index (ARI)
- Normalized Mutual Information (NMI)
- V-Measure