Matches in SemOpenAlex for { <https://semopenalex.org/work/W4287777281> ?p ?o ?g. }
Showing items 1 to 65 of
65
with 100 items per page.
- W4287777281 abstract "We present a new approach to solve the sparse approximation or best subset selection problem, namely find a $k$-sparse vector ${bf x}inmathbb{R}^d$ that minimizes the $ell_2$ residual $lVert A{bf x}-{bf y} rVert_2$. We consider a regularized approach, whereby this residual is penalized by the non-convex $textit{trimmed lasso}$, defined as the $ell_1$-norm of ${bf x}$ excluding its $k$ largest-magnitude entries. We prove that the trimmed lasso has several appealing theoretical properties, and in particular derive sparse recovery guarantees assuming successful optimization of the penalized objective. Next, we show empirically that directly optimizing this objective can be quite challenging. Instead, we propose a surrogate for the trimmed lasso, called the $textit{generalized soft-min}$. This penalty smoothly interpolates between the classical lasso and the trimmed lasso, while taking into account all possible $k$-sparse patterns. The generalized soft-min penalty involves summation over $binom{d}{k}$ terms, yet we derive a polynomial-time algorithm to compute it. This, in turn, yields a practical method for the original sparse approximation problem. Via simulations, we demonstrate its competitive performance compared to current state of the art." @default.
- W4287777281 created "2022-07-26" @default.
- W4287777281 creator A5008345510 @default.
- W4287777281 creator A5034942163 @default.
- W4287777281 creator A5058139643 @default.
- W4287777281 date "2020-05-18" @default.
- W4287777281 modified "2023-09-25" @default.
- W4287777281 title "The Trimmed Lasso: Sparse Recovery Guarantees and Practical Optimization by the Generalized Soft-Min Penalty" @default.
- W4287777281 doi "https://doi.org/10.48550/arxiv.2005.09021" @default.
- W4287777281 hasPublicationYear "2020" @default.
- W4287777281 type Work @default.
- W4287777281 citedByCount "0" @default.
- W4287777281 crossrefType "posted-content" @default.
- W4287777281 hasAuthorship W4287777281A5008345510 @default.
- W4287777281 hasAuthorship W4287777281A5034942163 @default.
- W4287777281 hasAuthorship W4287777281A5058139643 @default.
- W4287777281 hasBestOaLocation W42877772811 @default.
- W4287777281 hasConcept C112680207 @default.
- W4287777281 hasConcept C11413529 @default.
- W4287777281 hasConcept C114614502 @default.
- W4287777281 hasConcept C124851039 @default.
- W4287777281 hasConcept C126255220 @default.
- W4287777281 hasConcept C136764020 @default.
- W4287777281 hasConcept C155512373 @default.
- W4287777281 hasConcept C157972887 @default.
- W4287777281 hasConcept C17744445 @default.
- W4287777281 hasConcept C191795146 @default.
- W4287777281 hasConcept C199539241 @default.
- W4287777281 hasConcept C2524010 @default.
- W4287777281 hasConcept C28826006 @default.
- W4287777281 hasConcept C33923547 @default.
- W4287777281 hasConcept C37616216 @default.
- W4287777281 hasConcept C41008148 @default.
- W4287777281 hasConceptScore W4287777281C112680207 @default.
- W4287777281 hasConceptScore W4287777281C11413529 @default.
- W4287777281 hasConceptScore W4287777281C114614502 @default.
- W4287777281 hasConceptScore W4287777281C124851039 @default.
- W4287777281 hasConceptScore W4287777281C126255220 @default.
- W4287777281 hasConceptScore W4287777281C136764020 @default.
- W4287777281 hasConceptScore W4287777281C155512373 @default.
- W4287777281 hasConceptScore W4287777281C157972887 @default.
- W4287777281 hasConceptScore W4287777281C17744445 @default.
- W4287777281 hasConceptScore W4287777281C191795146 @default.
- W4287777281 hasConceptScore W4287777281C199539241 @default.
- W4287777281 hasConceptScore W4287777281C2524010 @default.
- W4287777281 hasConceptScore W4287777281C28826006 @default.
- W4287777281 hasConceptScore W4287777281C33923547 @default.
- W4287777281 hasConceptScore W4287777281C37616216 @default.
- W4287777281 hasConceptScore W4287777281C41008148 @default.
- W4287777281 hasLocation W42877772811 @default.
- W4287777281 hasOpenAccess W4287777281 @default.
- W4287777281 hasPrimaryLocation W42877772811 @default.
- W4287777281 hasRelatedWork W1490136338 @default.
- W4287777281 hasRelatedWork W2092550984 @default.
- W4287777281 hasRelatedWork W2372988593 @default.
- W4287777281 hasRelatedWork W2380702222 @default.
- W4287777281 hasRelatedWork W2897813401 @default.
- W4287777281 hasRelatedWork W3133920106 @default.
- W4287777281 hasRelatedWork W4205986199 @default.
- W4287777281 hasRelatedWork W4289367627 @default.
- W4287777281 hasRelatedWork W4293611323 @default.
- W4287777281 hasRelatedWork W2138044000 @default.
- W4287777281 isParatext "false" @default.
- W4287777281 isRetracted "false" @default.
- W4287777281 workType "article" @default.