For 1/f^1.5 signals, Gaussian, uniform, uniform-log, and Laplacian sparse feature sampling trace essentially the same test-error curve when indexed by empirical sampled-frequency standard deviation; scale matters more than distribution shape. · CiteArk
Not assessedPlan blockedFindingdistribution-shape-1d-15
For 1/f^1.5 signals, Gaussian, uniform, uniform-log, and Laplacian sparse feature sampling trace essentially the same test-error curve when indexed by empirical sampled-frequency standard deviation; scale matters more than distribution shape.
Source: paper:Figure 4(c), p. 7; §5 p. 6; Figure 13 p. 16
Reported and observed measurements
No structured measurement is attached to this Claim.
Assessments (0)
No immutable Assessment has been published for this Claim yet.
