Silhouette (clustering)¶
An internal cluster-validation score comparing each observation’s mean within-cluster dissimilarity with its smallest mean dissimilarity to another cluster.
Core Idea¶
For observation i, silhouette s(i)=(b(i)−a(i))/max{a(i),b(i)}, where a is mean dissimilarity within its assigned cluster and b is the best neighboring-cluster mean, producing values from −1 to 1. The score contrasts cohesion and separation at the point level; averaging or plotting values summarizes how assignments fit a chosen clustering and dissimilarity representation. The abstraction is therefore identified by a declared carrier, a transformation or constraint over that carrier, and an invariant that tells an analyst whether the named structure is genuinely present.
Scope of Application¶
Silhouette (clustering) belongs to cluster analysis and is useful where the analyst can specify the typed cluster analysis carrier, defining objects and relations, parameters, conventions, evidence, boundary cases, and comparison targets, then evaluate cluster assignments, dissimilarity metric, treatment of singleton clusters, neighboring-cluster rule, and aggregation are fixed before interpreting the bounded score. The scope is broad within that domain but bounded by the need for cluster assignments, dissimilarity metric, treatment of singleton clusters, neighboring-cluster rule, and aggregation are fixed before interpreting the bounded score. The entry records a descriptive analytical identity; practical use requires the governing domain's evidence, standards, and safety obligations.
Clarity¶
The abstraction clarifies a crowded vocabulary by making cluster assignments, dissimilarity metric, treatment of singleton clusters, neighboring-cluster rule, and aggregation are fixed before interpreting the bounded score the center of the account. A claim should name the carrier, the governing operation or relation, the applicable assumptions, and the recognition test. A bare label is insufficient because the name Silhouette (clustering) can be used for a formal identity, an implementation, or a neighboring result unless carrier and convention are stated.
Manages Complexity¶
Without the abstraction, an analyst must reason directly over many local details: the carrier roles, admissibility assumptions, competing conventions, derived invariants, boundary cases, and proof or validation obligations specific to Silhouette (clustering). Silhouette (clustering) compresses them into the roles in the structural signature. That compression permits comparison across instances without erasing the variables that determine validity. It also exposes which details may be varied safely and which are constitutive.
Abstract Reasoning¶
- Identify the carrier. State what the elements, states, objects, or observations are: the typed cluster analysis carrier, defining objects and relations, parameters, conventions, evidence, boundary cases, and comparison targets. Reject examples whose alleged carrier belongs to a different problem. 2. Lock the constitutive rule. Express cluster assignments, dissimilarity metric, treatment of singleton clusters, neighboring-cluster rule, and aggregation are fixed before interpreting the bounded score independently of one notation or implementation.
Knowledge Transfer¶
Knowledge transfers strongly among subfields of cluster analysis because they reuse the typed cluster analysis carrier, defining objects and relations, parameters, conventions, evidence, boundary cases, and comparison targets, The score contrasts cohesion and separation at the point level; averaging or plotting values summarizes how assignments fit a chosen clustering and dissimilarity representation., and type the carrier, state every parameter and convention in the definition, test that cluster assignments, dissimilarity metric, treatment of singleton clusters, neighboring-cluster rule, and aggregation are fixed before interpreting the bounded score, compare the nearest accepted identity, and report counterexamples, uncertainty, and limiting cases.
Relationships to Other Abstractions¶
Current abstraction Silhouette (clustering) Domain-specific
Parents (1) — more general patterns this builds on
-
Silhouette (clustering) is a kind of Comparison Prime
The proposed strict upward parent is
prime:comparison.
Hierarchy path (1) — routes to 1 parentless root
- Silhouette (clustering) → Comparison → Self Checking
Neighborhood in Abstraction Space¶
Silhouette (clustering) sits in a crowded region of the domain-specific corpus (36th percentile for distinctiveness): several abstractions share nearly its structure, so a description that fits it tends to fit its neighbors too.
Family — Cluster Validation & Sampling (8 abstractions)
Nearest neighbors
- Determining the number of clusters in a data set — 0.93
- Hopkins statistic — 0.91
- Rand index — 0.91
- K-means clustering — 0.91
- Davies–Bouldin index — 0.90
Computed from structural-signature embeddings · 2026-09-08