Skip to content
conftrace_
2026 ICML ICML 2026

Dimensionality Reduction with Point-Distribution Similarity Invariant

Existing dimensionality reduction methods all perform dimensionality reduction by preserving some invariant in the space before and after dimensionality reduction. This paper proposes a new dimensionality reduction invariant: preserving the invariant of the point-distributions similarity. We also design a linear and efficient method to achieve dimensionality reduction while preserving this invariant. We theoretically prove the feasibility of our method for dimensionality reduction. Furthermore, our results on benchmark datasets and single-cell expression data demonstrate the effectiveness and efficiency of the proposed method.

More from ICML 2026

All ICML 2026 papers →