GFTab, a semi-supervised tabular method with variable-specific corruptions and geodesic flow kernel similarity, reports the best F1 on about half of 21 mixed-variable benchmarks with sparse labels.
Grassmann angles between real or complex subspaces
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
The Grassmann angle improves upon similar angles between subspaces that measure volume contraction in orthogonal projections. It works in real or complex spaces, with important differences, and is asymmetric, what makes it more efficient when dimensions are distinct. It can be seen as an angle in Grassmann algebra, being related to its products and those of Clifford algebra, and gives the Fubini-Study metric on Grassmannians, an asymmetric metric on the full Grassmannian, and Hausdorff distances between full sub-Grassmannians. We give formulas for computing it in arbitrary bases, and identities for angles with certain families of subspaces, some of which are linked to real and complex Pythagorean theorems for volumes and quantum probabilities. Unusual features of the angle with an orthogonal complement, or the angle in complex spaces, are examined.
fields
cs.LG 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Geodesic Flow Kernels for Semi-Supervised Learning on Mixed-Variable Tabular Dataset
GFTab, a semi-supervised tabular method with variable-specific corruptions and geodesic flow kernel similarity, reports the best F1 on about half of 21 mixed-variable benchmarks with sparse labels.