Zobrazeno 1 - 5
of 5
pro vyhledávání: '"Michael Puthawala"'
Publikováno v:
Michael Puthawala
How can we design neural networks that allow for stable universal approximation of maps between topologically interesting manifolds? The answer is with a coordinate projection. Neural networks based on topological data analysis (TDA) use tools such a
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::c0d582637334edabcc5e30d841139f90
http://arxiv.org/abs/2210.00577
http://arxiv.org/abs/2210.00577
Publikováno v:
Michael Puthawala
University of Helsinki
University of Helsinki
Injectivity plays an important role in generative models where it enables inference; in inverse problems and compressed sensing with generative priors it is a precursor to well posedness. We establish sharp characterizations of injectivity of fully-c
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::d2f7a7b2c0d0804590b615c09aed4649
http://hdl.handle.net/10138/355819
http://hdl.handle.net/10138/355819
Publikováno v:
Michael Puthawala
We study approximation of probability measures supported on $n$-dimensional manifolds embedded in $\mathbb{R}^m$ by injective flows -- neural networks composed of invertible flows and injective layers. We show that in general, injective flows between
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::58d9e68481be845a10c6b553e9b75d16
http://arxiv.org/abs/2110.04227
http://arxiv.org/abs/2110.04227
Publikováno v:
Oberwolfach Reports; 2022, p745-789, 45p
Publikováno v:
Michael Puthawala
We investigate overdetermined linear inverse problems for which the forward operator may not be given accurately. We introduce a new tool called the structure, based on the Wasserstein distance, and propose the use of this to diagnose and remedy forw
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_dedup___::9bc7f005f0be3dedf108ac0ef4edc61d
http://arxiv.org/abs/1810.12993
http://arxiv.org/abs/1810.12993