SCSF accelerates eigenvalue dataset generation up to 3.5 times by grouping operators with similar spectra via truncated FFT sorting and reusing eigenpairs through Chebyshev subspace filters.
Following the approximation, we construct the matrix A, an N−1×N−1 tridiagonal matrix, crucial for the computations
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Accelerating Eigenvalue Dataset Generation via Chebyshev Subspace Filter
SCSF accelerates eigenvalue dataset generation up to 3.5 times by grouping operators with similar spectra via truncated FFT sorting and reusing eigenpairs through Chebyshev subspace filters.