4 papers
How Many Directions Must a Truncated Diffusion Sampler Retain? Matching Bounds Under Power-Law Spectra
Radmehr Karimian, Ali Mohades, Johannes Lederer
Diffusion samplers can reduce computation by generating selected spectral coordinates and filling the remaining directions with noise. How many directions must they retain? We stud…
Diffusion-based Denoising Beats Vanilla Score Matching in Parameter Estimation: A Theoretical Explanation
Benedikt Lütke Schwienhorst, Nadja Klein, Johannes Lederer
Score matching is an alternative to maximum likelihood estimation when the normalizing constant is unknown or too costly to evaluate. However, vanilla score matching has shown to b…
Regularized Learning for Fractional Brownian Motion via Path Signatures
Ali Mohaddes, Francesco Iafrate, Johannes Lederer
Fractional Brownian motion (fBm) extends classical Brownian motion by introducing dependence between increments, governed by the Hurst parameter . Unlike traditional Br…
Affine Invariance in Continuous-Domain Convolutional Neural Networks
Ali Mohaddes, Johannes Lederer
The notion of group invariance helps neural networks in recognizing patterns and features under geometric transformations. Group convolutional neural networks enhance traditional c…