The paper reviews a three-axis evaluation framework (stability, cluster quality, topological shape preservation) for Mapper algorithms, analyzes variants on synthetic and UCI Digits data, and finds the axes often conflict with no single variant optimal across all.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.AT 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
A Three Axis Evaluation Framework for Mapper Algorithms
The paper reviews a three-axis evaluation framework (stability, cluster quality, topological shape preservation) for Mapper algorithms, analyzes variants on synthetic and UCI Digits data, and finds the axes often conflict with no single variant optimal across all.