Matches in SemOpenAlex for { <https://semopenalex.org/work/W4310282761> ?p ?o ?g. }
Showing items 1 to 71 of
71
with 100 items per page.
- W4310282761 abstract "It is well-known that any matrix $A$ has an LU decomposition. Less well-known is the fact that it has a 'Toeplitz decomposition' $A = T_1 T_2 cdots T_r$ where $T_i$'s are Toeplitz matrices. We will prove that any continuous function $f : mathbb{R}^n to mathbb{R}^m$ has an approximation to arbitrary accuracy by a neural network that takes the form $L_1 sigma_1 U_1 sigma_2 L_2 sigma_3 U_2 cdots L_r sigma_{2r-1} U_r$, i.e., where the weight matrices alternate between lower and upper triangular matrices, $sigma_i(x) := sigma(x - b_i)$ for some bias vector $b_i$, and the activation $sigma$ may be chosen to be essentially any uniformly continuous nonpolynomial function. The same result also holds with Toeplitz matrices, i.e., $f approx T_1 sigma_1 T_2 sigma_2 cdots sigma_{r-1} T_r$ to arbitrary accuracy, and likewise for Hankel matrices. A consequence of our Toeplitz result is a fixed-width universal approximation theorem for convolutional neural networks, which so far have only arbitrary width versions. Since our results apply in particular to the case when $f$ is a general neural network, we may regard them as LU and Toeplitz decompositions of a neural network. The practical implication of our results is that one may vastly reduce the number of weight parameters in a neural network without sacrificing its power of universal approximation. We will present several experiments on real data sets to show that imposing such structures on the weight matrices sharply reduces the number of training parameters with almost no noticeable effect on test accuracy." @default.
- W4310282761 created "2022-11-30" @default.
- W4310282761 creator A5025152292 @default.
- W4310282761 creator A5050794507 @default.
- W4310282761 creator A5086634045 @default.
- W4310282761 date "2022-11-25" @default.
- W4310282761 modified "2023-09-27" @default.
- W4310282761 title "LU decomposition and Toeplitz decomposition of a neural network" @default.
- W4310282761 doi "https://doi.org/10.48550/arxiv.2211.13935" @default.
- W4310282761 hasPublicationYear "2022" @default.
- W4310282761 type Work @default.
- W4310282761 citedByCount "0" @default.
- W4310282761 crossrefType "posted-content" @default.
- W4310282761 hasAuthorship W4310282761A5025152292 @default.
- W4310282761 hasAuthorship W4310282761A5050794507 @default.
- W4310282761 hasAuthorship W4310282761A5086634045 @default.
- W4310282761 hasBestOaLocation W43102827611 @default.
- W4310282761 hasConcept C106487976 @default.
- W4310282761 hasConcept C114614502 @default.
- W4310282761 hasConcept C118615104 @default.
- W4310282761 hasConcept C121332964 @default.
- W4310282761 hasConcept C124681953 @default.
- W4310282761 hasConcept C14036430 @default.
- W4310282761 hasConcept C147710293 @default.
- W4310282761 hasConcept C154945302 @default.
- W4310282761 hasConcept C159985019 @default.
- W4310282761 hasConcept C18903297 @default.
- W4310282761 hasConcept C192562407 @default.
- W4310282761 hasConcept C202444582 @default.
- W4310282761 hasConcept C2778049214 @default.
- W4310282761 hasConcept C33923547 @default.
- W4310282761 hasConcept C41008148 @default.
- W4310282761 hasConcept C50644808 @default.
- W4310282761 hasConcept C62520636 @default.
- W4310282761 hasConcept C78458016 @default.
- W4310282761 hasConcept C86803240 @default.
- W4310282761 hasConceptScore W4310282761C106487976 @default.
- W4310282761 hasConceptScore W4310282761C114614502 @default.
- W4310282761 hasConceptScore W4310282761C118615104 @default.
- W4310282761 hasConceptScore W4310282761C121332964 @default.
- W4310282761 hasConceptScore W4310282761C124681953 @default.
- W4310282761 hasConceptScore W4310282761C14036430 @default.
- W4310282761 hasConceptScore W4310282761C147710293 @default.
- W4310282761 hasConceptScore W4310282761C154945302 @default.
- W4310282761 hasConceptScore W4310282761C159985019 @default.
- W4310282761 hasConceptScore W4310282761C18903297 @default.
- W4310282761 hasConceptScore W4310282761C192562407 @default.
- W4310282761 hasConceptScore W4310282761C202444582 @default.
- W4310282761 hasConceptScore W4310282761C2778049214 @default.
- W4310282761 hasConceptScore W4310282761C33923547 @default.
- W4310282761 hasConceptScore W4310282761C41008148 @default.
- W4310282761 hasConceptScore W4310282761C50644808 @default.
- W4310282761 hasConceptScore W4310282761C62520636 @default.
- W4310282761 hasConceptScore W4310282761C78458016 @default.
- W4310282761 hasConceptScore W4310282761C86803240 @default.
- W4310282761 hasLocation W43102827611 @default.
- W4310282761 hasOpenAccess W4310282761 @default.
- W4310282761 hasPrimaryLocation W43102827611 @default.
- W4310282761 hasRelatedWork W1810226790 @default.
- W4310282761 hasRelatedWork W1978042415 @default.
- W4310282761 hasRelatedWork W1998305433 @default.
- W4310282761 hasRelatedWork W2043721180 @default.
- W4310282761 hasRelatedWork W2115605241 @default.
- W4310282761 hasRelatedWork W2185733537 @default.
- W4310282761 hasRelatedWork W2354178226 @default.
- W4310282761 hasRelatedWork W2478221555 @default.
- W4310282761 hasRelatedWork W4233303626 @default.
- W4310282761 hasRelatedWork W49739194 @default.
- W4310282761 isParatext "false" @default.
- W4310282761 isRetracted "false" @default.
- W4310282761 workType "article" @default.