INFS4203/7203 Data Mining Tutorial Week 6 Semester 2, 2023
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
INFS4203/7203 Data Mining
Tutorial Week 6 - Density-based Clustering and Hierarchical-based Clustering
Semester 2, 2023
Question 1/2: Density-based Clustering
The following image shows the distribution of 8 two-dimensional points.
A circle with radius Eps is drawn around each of the eight points: A, B, C, D, E, F, G, H.
1. If we set MinPts = 3, discuss the category of these eight points according to the DBSCAN algorithm, i.e., which of these points are core/border/noise points.
2. If we set MinPts = 3, how does a DBSCAN clustering work on these points?
3. If we set MinPts = 4, what are the core/border/noise points? What is the result for DBSCAN? Compare the results when MinPts = 3 and MinPts = 4.
4. (Extra) Prove that in DBSCAN, for a fixed MinPts value and two neighborhood thresholds ϵ 1 < ϵ2 , a cluster C with respect to ϵ1 and MinPts must be a subset of a cluster C′ with respect to ϵ2 and MinPts.
Question 2/2: Hierarchical-based Clustering
Suppose the data mining task is to use agglomerative clustering to group measurements of the variable weight.
1. Commute the step-by-step agglomerative grouping using: Single linkage and Complete linkage with Manhattan distance.
weight = {40, 63, 54, 11, 33, 29}
2. Commute the step-by-step agglomerative grouping using: Average linkage with Manhattan distance.
weight = {40, 63, 54, 11, 33, 28}
3. According to the Agglomerative method, what is the cluster result (how to assign the points to each cluster) if we need two clusters with respect to single linkage strategy? How about three clusters?
4. (Extra) Consider the following four faces shown. Darkness or number of dots represents density. Lines are used only to distinguish regions and do not represent points.
For each figure, could you use single link to find the patterns represented by the nose, eyes, and mouth? Explain. Patterns are defined as clusters, e.g. we have a pattern for the left eye if there is a cluster contains points from left eye only.
2023-09-06
Density-based Clustering and Hierarchical-based Clustering