Clustering — k-means finds groups even when there are none
k-means and friends partition data into groups, but they will always return groups, whether or not the data has any — so the hard part is validating that a cluster means something.
Not yet scoped.
Why
The PoC clusters random noise to show spurious structure, then uses silhouette and stability checks to separate a real segmentation from an artifact of the chosen k.
How it works
Not yet built.
Where it lands in Jayverse
Personas: validate before treating a cluster as a real segment. Wallet or collector clustering must ship with a silhouette or stability check against the chosen k, not just a reported cluster count.
Number: don't publish a "market regime" without the same check. Any reading that clusters market or wallet data needs a validation section, since an unvalidated cluster reads as insight but may just be k-means finding structure in noise.
Key expressions
Words and phrases from this page worth keeping, with the Korean meaning and the sentence they come from.
Expression
뜻 · 쓰이는 자리
spurious structure
가짜(허위) 구조 · 실제로는 없는 패턴인데 군집화 알고리즘이 억지로 만들어낸 것 · "clusters random noise to show spurious structure"
silhouette check
실루엣 검증 · 각 데이터가 자기 군집에 얼마나 잘 맞는지 재는 지표 검사 · "uses silhouette and stability checks to separate"
stability check
안정성 검증 · 조건을 조금씩 바꿔도 같은 군집이 재현되는지 확인하는 절차 · "and stability checks to separate a real"
artifact of
~로 인해 생긴 인위적 산물 · 실제 신호가 아니라 설정값(k) 때문에 생긴 결과라는 뜻 · "an artifact of the chosen k"
partition into
~로 나누다, 분할하다 · 데이터를 여러 그룹으로 쪼갠다는 뜻 · "partition data into groups"
k-means
k-평균 군집화 · 데이터를 k개의 중심점 기준으로 나누는 대표적 클러스터링 알고리즘, 데이터에 구조가 없어도 항상 군집을 반환함. "k-means and friends partition data into groups"