Matches in SemOpenAlex for { <https://semopenalex.org/work/W3167942312> ?p ?o ?g. }
- W3167942312 abstract "Multi-modal distributions are commonly used to model clustered data in statistical learning tasks. In this paper, we consider the Mixed Linear Regression (MLR) problem. We propose an optimal transport-based framework for MLR problems, Wasserstein Mixed Linear Regression (WMLR), which minimizes the Wasserstein distance between the learned and target mixture regression models. Through a model-based duality analysis, WMLR reduces the underlying MLR task to a nonconvex-concave minimax optimization problem, which can be provably solved to find a minimax stationary point by the Gradient Descent Ascent (GDA) algorithm. In the special case of mixtures of two linear regression models, we show that WMLR enjoys global convergence and generalization guarantees. We prove that WMLR's sample complexity grows linearly with the dimension of data. Finally, we discuss the application of WMLR to the federated learning task where the training samples are collected by multiple agents in a network. Unlike the Expectation Maximization algorithm, WMLR directly extends to the distributed, federated learning setting. We support our theoretical results through several numerical experiments, which highlight our framework's ability to handle the federated learning setting with mixture models." @default.
- W3167942312 created "2021-06-22" @default.
- W3167942312 creator A5005913897 @default.
- W3167942312 creator A5016741616 @default.
- W3167942312 creator A5017160178 @default.
- W3167942312 creator A5033517161 @default.
- W3167942312 creator A5067307504 @default.
- W3167942312 date "2021-06-14" @default.
- W3167942312 modified "2023-10-16" @default.
- W3167942312 title "A Wasserstein Minimax Framework for Mixed Linear Regression" @default.
- W3167942312 cites W1486632395 @default.
- W3167942312 cites W1974648697 @default.
- W3167942312 cites W2025653905 @default.
- W3167942312 cites W2099471712 @default.
- W3167942312 cites W2124428910 @default.
- W3167942312 cites W2138967244 @default.
- W3167942312 cites W2550703763 @default.
- W3167942312 cites W2577946330 @default.
- W3167942312 cites W2593729559 @default.
- W3167942312 cites W2739748921 @default.
- W3167942312 cites W2787248994 @default.
- W3167942312 cites W2788819334 @default.
- W3167942312 cites W2803117117 @default.
- W3167942312 cites W2890485411 @default.
- W3167942312 cites W2962688627 @default.
- W3167942312 cites W2962737134 @default.
- W3167942312 cites W2963506485 @default.
- W3167942312 cites W2964207716 @default.
- W3167942312 cites W2965628250 @default.
- W3167942312 cites W3007548213 @default.
- W3167942312 cites W3022507773 @default.
- W3167942312 cites W3034163621 @default.
- W3167942312 cites W3034666238 @default.
- W3167942312 cites W3035012305 @default.
- W3167942312 cites W3036353678 @default.
- W3167942312 cites W3037414083 @default.
- W3167942312 cites W3039159133 @default.
- W3167942312 cites W3080934299 @default.
- W3167942312 cites W3099314130 @default.
- W3167942312 cites W3104015155 @default.
- W3167942312 cites W3129603732 @default.
- W3167942312 hasPublicationYear "2021" @default.
- W3167942312 type Work @default.
- W3167942312 sameAs 3167942312 @default.
- W3167942312 citedByCount "0" @default.
- W3167942312 crossrefType "posted-content" @default.
- W3167942312 hasAuthorship W3167942312A5005913897 @default.
- W3167942312 hasAuthorship W3167942312A5016741616 @default.
- W3167942312 hasAuthorship W3167942312A5017160178 @default.
- W3167942312 hasAuthorship W3167942312A5033517161 @default.
- W3167942312 hasAuthorship W3167942312A5067307504 @default.
- W3167942312 hasConcept C114614502 @default.
- W3167942312 hasConcept C119857082 @default.
- W3167942312 hasConcept C126255220 @default.
- W3167942312 hasConcept C134306372 @default.
- W3167942312 hasConcept C149728462 @default.
- W3167942312 hasConcept C153258448 @default.
- W3167942312 hasConcept C154945302 @default.
- W3167942312 hasConcept C162324750 @default.
- W3167942312 hasConcept C177148314 @default.
- W3167942312 hasConcept C2777303404 @default.
- W3167942312 hasConcept C33676613 @default.
- W3167942312 hasConcept C33923547 @default.
- W3167942312 hasConcept C41008148 @default.
- W3167942312 hasConcept C41045048 @default.
- W3167942312 hasConcept C48921125 @default.
- W3167942312 hasConcept C50522688 @default.
- W3167942312 hasConcept C50644808 @default.
- W3167942312 hasConceptScore W3167942312C114614502 @default.
- W3167942312 hasConceptScore W3167942312C119857082 @default.
- W3167942312 hasConceptScore W3167942312C126255220 @default.
- W3167942312 hasConceptScore W3167942312C134306372 @default.
- W3167942312 hasConceptScore W3167942312C149728462 @default.
- W3167942312 hasConceptScore W3167942312C153258448 @default.
- W3167942312 hasConceptScore W3167942312C154945302 @default.
- W3167942312 hasConceptScore W3167942312C162324750 @default.
- W3167942312 hasConceptScore W3167942312C177148314 @default.
- W3167942312 hasConceptScore W3167942312C2777303404 @default.
- W3167942312 hasConceptScore W3167942312C33676613 @default.
- W3167942312 hasConceptScore W3167942312C33923547 @default.
- W3167942312 hasConceptScore W3167942312C41008148 @default.
- W3167942312 hasConceptScore W3167942312C41045048 @default.
- W3167942312 hasConceptScore W3167942312C48921125 @default.
- W3167942312 hasConceptScore W3167942312C50522688 @default.
- W3167942312 hasConceptScore W3167942312C50644808 @default.
- W3167942312 hasOpenAccess W3167942312 @default.
- W3167942312 hasRelatedWork W1966270882 @default.
- W3167942312 hasRelatedWork W2107151623 @default.
- W3167942312 hasRelatedWork W2169603879 @default.
- W3167942312 hasRelatedWork W2770847996 @default.
- W3167942312 hasRelatedWork W2790379555 @default.
- W3167942312 hasRelatedWork W2804016305 @default.
- W3167942312 hasRelatedWork W2805564340 @default.
- W3167942312 hasRelatedWork W2889431799 @default.
- W3167942312 hasRelatedWork W2920218480 @default.
- W3167942312 hasRelatedWork W2951140320 @default.
- W3167942312 hasRelatedWork W2954402037 @default.
- W3167942312 hasRelatedWork W2963581809 @default.
- W3167942312 hasRelatedWork W2975364676 @default.
- W3167942312 hasRelatedWork W3087642357 @default.