What is the main concept behind hierarchical clustering?

Question

BlackTom AI · Accepted Answer

Hierarchical clustering builds a tree-like hierarchy of clusters; the most common form is agglomerative (bottom-up), which is precisely described by the correct option.

Relevant concepts:
- Agglomerative vs divisive: agglomerative starts with each observation as its own cluster and merges them; divisive starts with one cluster and splits it. 
- Distance matrix: pairwise distances between points guide which clusters to merge. 
- Linkage criteria: single, complete, average, or Ward linkage determine how to compute distance between clusters after a merge. 
- Dendrogram: a tree diagram showing the order and distance of merges; cutting the dendrogram at different heights yields different clusterings.

Step-by-step reasoning (complete process):
1. Initialize: each data point is its own cluster. Example points A, B, C, D.
2. Compute pairwise distances (distance matrix). Suppose distances: AB=1, AC=4, AD=5, BC=3, BD=6, CD=2.
3. Find the smallest distance: AB=1, so merge A and B into cluster {A,B}.
4. Update distances between {A,B} and other points using a linkage rule. With single linkage, d({A,B},C)=min(d(A,C),d(B,C)) = min(4,3)=3; d({A,B},D)=min(5,6)=5.
5. From updated distances (to C:3, to D:5, CD=2), smallest is CD=2 → merge C and D → cluster {C,D}.
6. Now merge the two clusters {A,B} and {C,D} based on their inter-cluster distance (e.g., using chosen linkage) resulting in a single cluster containing all points.
7. The process stops when all points are merged into one cluster; intermediate states form a dendrogram.

Why the correct answer is correct:
- It explicitly describes the agglomerative hierarchical method: start with singleton clusters and iteratively combine nearest clusters until one cluster remains.

Why other options are incorrect:
- Option 1 is vague and inaccurate (no standard “identify most valuable clusters by gradually building them from a hierarchy of points”).
- Option 2 describes k-means or partition-based clustering where the number of clusters is fixed beforehand.
- Option 4 suggests random selection and iterative assignment, which fits methods like k-means initialization or random clustering, not hierarchical clustering.

Conclusion: The correct answer is "We begin with each point as their own cluster and combine nearby points into clusters until they are all one cluster," which matches the authoritative answer provided.

What is the main concept behind hierarchical clustering?单项选择题

类似问题

Which of the following is an advantage of hierarchical clustering over k-Means clustering?

What is a dendrogram?

If a dendrogram shows two main branches that merge at a very high distance (height) on the vertical axis, what does this suggest?

Which rule defines the "Complete Linkage" method in Hierarchical clustering?

In an agglomerative hierarchical clustering dendrogram, how do you determine the number of clusters by "cutting" the tree?

Which of the following is finally produced by hierarchical clustering?

Hierarchical clustering differs from K-means clustering in that:

更多留学生实用工具

智能学习助手

风格化写作助手

论文查重助手

文献引用助手

课堂转译助手

课堂笔记助手

Quiz搜索助手

学校历年真题

智能刷题助手

智能匹配练习

希望你的学习变得更简单