Spelling suggestions: "subject:"locality constraints"" "subject:"vocality constraints""
1 |
Idiomatic Root Merge in Modern Hebrew blendsPham, Mike January 2011 (has links)
In this paper I use the Distributional Morphology framework and semantic Locality Constraints proposed by Arad (2003) to look at category assignments of blends in Modern Hebrew, as well as blends, compounds and idioms in English where relevant. Bat-El (1996) provides an explicit phonological analysis of Modern Hebrew blends, and argues against any morphological process at play in blend formation. I argue, however, that blends and compounds must be accounted for within morphology due to category assignments. I first demonstrate that blends are unquestionably formed by blending fully inflected words rather than roots, and then subsequently reject an analysis that accounts for weakened Locality Constraints by proposing the formation of a new root. Instead, I propose a hypothesis of Idiomatic Root Merge where a root can be an n-place predicate that selects at least an XP sister and a category head. This proposal also entails that there is a structural difference between two surface-similar phrases that have respectively literal and idiomatic meanings.
|
2 |
Graph Bandits : Multi-Armed Bandits with Locality Constraints / Grafbanditer : Flerarmade banditer med lokala restriktionerJohansson, Kasper January 2022 (has links)
Multi-armed bandits (MABs) have been studied extensively in the literature and have applications in a wealth of domains, including recommendation systems, dynamic pricing, and investment management. On the one hand, the current MAB literature largely seems to focus on the setting where each arm is available to play at each time step, and ignores how agents move between the arms. On the other hand, there is work that takes the movement between arms into account, but this work models the problem as a Markov decision process and applies generic reinforcement learning (RL) algorithms, like Q-learning. This thesis examines an extension of the MAB problem to a setting where the set of available arms at each round depends on which arm was played in the previous round. In this formulation the arms are nodes in a graph, and arms that can be played successively are connected via edges. We denote this the Graph Bandit (GB) problem. We show that under certain conditions the optimal action is governed by a stationary policy. Furthermore, we develop an algorithm that leverages the graphical structure of the problem to find this policy when the reward distributions are perfectly known, and denote this algorithm the Q-graph. When the reward distributions are unknown, we show how to leverage the Qgraph algorithm together with standard sampling algorithms like Thompson sampling and upper confidence bound to create an online learning algorithm that provably achieves logarithmic regret. Finally, this regret-bound is supported in numerical simulations, and it is illustrated how the proposed Q-graph algorithm outperforms generic algorithms from the MAB and RL communities. / Flerarmade banditer (FAB) har studerats omfattande i litteraturen och har applikationer inom en mängd domäner, såsom rekommendationssystem, dynamisk prissättning och finans. Å ena sidan verkar det som at en stor del av litteraturen fokuserar på situationen där alla armar är tillgängliga att spela vid varje tidssteg och ignorerar hur agenten rör sig mellan armarna. Å andra sidan finns det arbete som tar till hänsyn hur agenten rör sig mellan armarna men det arbetet modellerar systemet som en Markovprocess och använder sig av generiska inlärningsmetoder, såsom Q-learning. Den här uppsatsen undersöker en utvidgning av FAB-problemet till en situation där mängden tillgänliga armar vid varje runda beror på vilken arm som spelades i den föregående rundan. I denna formulering är armarna noder i en graf och armar som kan spelas i på varandra följande rundor är anslutna via kanter. Vi kallar det här problemt Grafbanditen. Vi visar att under vissa förutsättningar bestäms det optimala aggerandet av en stationär policy. Vi utvecklar också en algoritm som utnyttjar den grafiska strukturen i problemet för att beräkna denna policy när distributionerna hos alla armar är kända. Denna algoritm får namnet Q-grafen. När distributionerna är okända visar vi hur Q-grafen kan användas tillsammans med Thompson sampling eller upper confidence bound-metoder för att skapa en online inlärningsalgoritm som bevisligen uppnår logaritmisk regret. Slutligen stöds de teoretiska resultaten via numeriska simuleringar som illustrerar att Q-grafen är överlägsen många generiska inlärningsalgoritmer.
|
3 |
Restrições de extração de argumentos e adjuntos de nome no português brasileiro / Constraints on extration of arguments and nominal adjuncts in brazilian portugueseSedrins, Adeilson Pinheiro 19 July 2009 (has links)
The main goal of this thesis is to analyze the syntactic structure of determiner phrases in Brazilian Portuguese (BP), under the Principles and Parameters approach in its minimalist version (CHOMSKY, 1995 and subsequent works), focusing on the extraction of arguments and adjuncts introduced by de (of) preposition (de-phrases). The analysis to be developed takes into account Abney s (1987) DP hypothesis and the theoretical perspective that assumes the nominal construction as the reflex of the architecture of the clause structure. The de-phrases are going to be analyzed into two different groups: those which are true genitive constructions (arguments) and those which we are going to refer to as typical adjuncts. As we will make clear throughout the thesis, the group of genitives is generated in argument positions and the second group is generated in adjunction configuration. In what regards the restrictions imposed on extraction of de-phrases, we are going to propose that it can be captured by the unique constraint of Locality, as discussed in Manzini (1994). We are also going to argue that in BP the locus of phi-features (at least in what concerns number features) is the head D (MAGALHÃES, 2004; COSTA & FIGUEIREDOSILVA, 2006) and, once D is the locus for number in this language, this head is responsible for the Case properties of genitives (following similar analysis showed in AVELAR, 2006). Another crucial point of the proposal is the projection of a category FP, above DP, in BP nominal constructions, which plays similar role of the projection TopP proposed in Haegeman (2004). [Spec, FP] is an escape hatch position, not [Spec, DP], the last one an A position, related to Case properties. In relation to adjunction phenomenon, we will assume Hornstein & Nunes s (2008) proposal that adjunction occurs without obligatory labeling, in a similar way that Avelar (2006) proposes adjunction in DPs of BP. Instead of being adjoined in the DP periphery, as stated in Avelar, true nominal adjuncts are adjoined in a projection below DP, fact that explains why typical adjuncts can t be extracted in this language. / Conselho Nacional de Desenvolvimento Científico e Tecnológico / O principal objetivo desta tese é analisar a estrutura sintática dos sintagmas de determinante no Português Brasileiro (PB), dentro do quadro teórico da Teoria de Princípios e Parâmetros, na sua versão Minimalista (CHOMSKY, 1995 e trabalhos subseqüentes), centrando-se no fenômeno da extração de argumentos e adjuntos de nome introduzidos pela preposição de (de-phrases), como um fenômeno submetido a restrições de ordem estrutural. A análise a ser apresentada assume a hipótese DP como proposta em Abney (1987), bem como a perspectiva teórica que tem tratado a arquitetura do sintagma nominal como uma arquitetura que reflete a da sentença. Os de-phrases em contextos nominais serão analisados separadamente em dois grupos: o grupo dos sintagmas genuinamente genitivos e que são manipulados na sintaxe como verdadeiros DPs, aos quais iremos nos referir como sintagmas argumentais, e o grupo a que iremos nos referir como típicos adjuntos adnominais. Conforme iremos mostrar ao longo desta tese, enquanto que os sintagmas do primeiro grupo são licenciados em posições argumentais, o segundo grupo é licenciado em configuração de adjunção. No que concerne às restrições impostas para a extração dos de-phrases, iremos propor que as restrições a serem verificadas podem ser capturadas unicamente pela noção da condição de Localidade, como proposto em Manzini (1994). Ainda, iremos argumentar que, no PB, os traços de concordância (pelo menos no que se refere à concordância de número) parecem estar localizados em D (MAGALHÃES, 2004; COSTA & FIGUEIREDO-SILVA, 2006) e, uma vez que D é lócus de número no PB, esta categoria licencia genitivos em relação a Caso, idéia já presente em Avelar (2006). Em termos de estrutura do DP, outro ponto crucial da análise é a proposta de projeção de uma categoria funcional acima da projeção DP, a qual denominaremos FP, nos moldes em que TopP é uma categoria proposta em Haegeman (2004) no domínio nominal. Conforme iremos discutir, a posição [Spec, FP] é a posição de válvula de escape para de-phrases no PB e não a posição [Spec, DP], esta última, uma posição A, relacionada a propriedades de Caso (genitivo). No que tange ao fenômeno da adjunção dentro de sintagmas de determinante, iremos assumir a proposta de Hornstein & Nunes (2008) de que a adjunção é uma operação que ocorre sem rotulação obrigatória, seguindo a proposta de adjunção de de-phrases no sintagma nominal como apresentada em Avelar (2006), para o PB. A diferença em relação à proposta de Avelar é a de que em vez de serem licenciados na periferia da construção nominal, os típicos adjuntos adnominais são gerados abaixo da projeção DP, o que explica a resistência desses constituintes para extração.
|
4 |
Serbo-Croatian Word Order: A Logical ApproachMihalicek, Vedrana 18 December 2012 (has links)
No description available.
|
Page generated in 0.0964 seconds