Matches in SemOpenAlex for { <https://semopenalex.org/work/W2920084148> ?p ?o ?g. }
Showing items 1 to 93 of
93
with 100 items per page.
- W2920084148 abstract "High-dimensional statistical learning (HDSL) has been widely applied in data analysis, operations research, and stochastic optimization. Despite the availability of multiple theoretical frameworks, most HDSL theories stipulate the following two conditions, which are sometimes overly critical: (a) the sparsity, and (b) the restricted strong convexity (RSC). This paper generalizes both conditions via the use of the folded concave penalty (FCP); we show that, for an M-estimation problem where (i) the (conventional) sparsity is relaxed into the approximate sparsity and (ii) the RSC is completely absent, the FCP-based regularization leads to poly-logarithmic sample complexity: the size of the training data is only required to be poly-logarithmic in the problem dimensionality. This finding allows us to further understand two important paradigms much less discussed formerly: the high-dimensional nonsmooth learning and the (deep) neural networks (NN). For both problems, we show that the poly-logarithmic sample complexity can be maintained. Furthermore, via integrating the NN with the FCP, the excess risk of a stationary point to the training formulation for the NN is strictly monotonic with respect to the solution's suboptimality gap, providing the first theoretical evidence for the empirically observed consistency between the generalization performance and the optimization quality in training an NN." @default.
- W2920084148 created "2019-03-11" @default.
- W2920084148 creator A5041526408 @default.
- W2920084148 creator A5087945394 @default.
- W2920084148 date "2019-03-02" @default.
- W2920084148 modified "2023-09-27" @default.
- W2920084148 title "High-Dimensional Learning under Approximate Sparsity: A Unifying Framework for Nonsmooth Learning and Regularized Neural Networks" @default.
- W2920084148 cites W1618978521 @default.
- W2920084148 cites W2384495648 @default.
- W2920084148 cites W2513671774 @default.
- W2920084148 cites W2807299122 @default.
- W2920084148 cites W2889756710 @default.
- W2920084148 cites W2962698540 @default.
- W2920084148 cites W2963452074 @default.
- W2920084148 cites W2963828549 @default.
- W2920084148 cites W2990200055 @default.
- W2920084148 cites W3119586787 @default.
- W2920084148 cites W3137695714 @default.
- W2920084148 hasPublicationYear "2019" @default.
- W2920084148 type Work @default.
- W2920084148 sameAs 2920084148 @default.
- W2920084148 citedByCount "1" @default.
- W2920084148 countsByYear W29200841482020 @default.
- W2920084148 crossrefType "posted-content" @default.
- W2920084148 hasAuthorship W2920084148A5041526408 @default.
- W2920084148 hasAuthorship W2920084148A5087945394 @default.
- W2920084148 hasConcept C106159729 @default.
- W2920084148 hasConcept C111030470 @default.
- W2920084148 hasConcept C126255220 @default.
- W2920084148 hasConcept C134306372 @default.
- W2920084148 hasConcept C154945302 @default.
- W2920084148 hasConcept C162324750 @default.
- W2920084148 hasConcept C177148314 @default.
- W2920084148 hasConcept C185592680 @default.
- W2920084148 hasConcept C198531522 @default.
- W2920084148 hasConcept C2776135515 @default.
- W2920084148 hasConcept C2776436953 @default.
- W2920084148 hasConcept C28826006 @default.
- W2920084148 hasConcept C2984842247 @default.
- W2920084148 hasConcept C33923547 @default.
- W2920084148 hasConcept C39927690 @default.
- W2920084148 hasConcept C41008148 @default.
- W2920084148 hasConcept C43617362 @default.
- W2920084148 hasConcept C50644808 @default.
- W2920084148 hasConcept C72134830 @default.
- W2920084148 hasConcept C72169020 @default.
- W2920084148 hasConceptScore W2920084148C106159729 @default.
- W2920084148 hasConceptScore W2920084148C111030470 @default.
- W2920084148 hasConceptScore W2920084148C126255220 @default.
- W2920084148 hasConceptScore W2920084148C134306372 @default.
- W2920084148 hasConceptScore W2920084148C154945302 @default.
- W2920084148 hasConceptScore W2920084148C162324750 @default.
- W2920084148 hasConceptScore W2920084148C177148314 @default.
- W2920084148 hasConceptScore W2920084148C185592680 @default.
- W2920084148 hasConceptScore W2920084148C198531522 @default.
- W2920084148 hasConceptScore W2920084148C2776135515 @default.
- W2920084148 hasConceptScore W2920084148C2776436953 @default.
- W2920084148 hasConceptScore W2920084148C28826006 @default.
- W2920084148 hasConceptScore W2920084148C2984842247 @default.
- W2920084148 hasConceptScore W2920084148C33923547 @default.
- W2920084148 hasConceptScore W2920084148C39927690 @default.
- W2920084148 hasConceptScore W2920084148C41008148 @default.
- W2920084148 hasConceptScore W2920084148C43617362 @default.
- W2920084148 hasConceptScore W2920084148C50644808 @default.
- W2920084148 hasConceptScore W2920084148C72134830 @default.
- W2920084148 hasConceptScore W2920084148C72169020 @default.
- W2920084148 hasLocation W29200841481 @default.
- W2920084148 hasOpenAccess W2920084148 @default.
- W2920084148 hasPrimaryLocation W29200841481 @default.
- W2920084148 hasRelatedWork W1543618960 @default.
- W2920084148 hasRelatedWork W2124140601 @default.
- W2920084148 hasRelatedWork W2132054311 @default.
- W2920084148 hasRelatedWork W2233806934 @default.
- W2920084148 hasRelatedWork W2382022742 @default.
- W2920084148 hasRelatedWork W2414029883 @default.
- W2920084148 hasRelatedWork W2502891885 @default.
- W2920084148 hasRelatedWork W2767727533 @default.
- W2920084148 hasRelatedWork W2789587413 @default.
- W2920084148 hasRelatedWork W2911498438 @default.
- W2920084148 hasRelatedWork W2950038553 @default.
- W2920084148 hasRelatedWork W2953149562 @default.
- W2920084148 hasRelatedWork W2962794632 @default.
- W2920084148 hasRelatedWork W2962835214 @default.
- W2920084148 hasRelatedWork W2999664547 @default.
- W2920084148 hasRelatedWork W3013021751 @default.
- W2920084148 hasRelatedWork W3110884700 @default.
- W2920084148 hasRelatedWork W3114692508 @default.
- W2920084148 hasRelatedWork W3207554517 @default.
- W2920084148 hasRelatedWork W3208461153 @default.
- W2920084148 isParatext "false" @default.
- W2920084148 isRetracted "false" @default.
- W2920084148 magId "2920084148" @default.
- W2920084148 workType "article" @default.