<p>The last two decades witnessed an increasing emphasis on reverse logistics (RL). Our thesis attempts to investigate two research problems in RL and explore game theoretic applications in this field.</p> <p>In Chapter 1, we introduce SCM, RL, relevant game theoretic applications, and the organizational structure of this thesis.</p> <p>In Chapter 2, we address a newsvendor problem with resalable returns. We develop a basic model with order quantity as the single decision variable and conduct concavity analysis. We also develop a general model in which the retailer determines both order quantity and two inter-period inventory thresholds. We use simulation to investigate the timing effect of both customer demands and returns on the retailer's decision making.</p> <p>In Chapter 3, we explore the application of game theoretic models with incomplete information in inventory management. Games with incomplete information may provide a more realistic modeling framework. We hope this exposition be helpful to researchers interested in applying game theoretic models and computing equilibriums in their specific problems in SCM and RL.</p> <p>In Chapter 4 we consider a remanufacturing competition problem between an original equipment manufacturer (OEM) and a pure remanufacturer (REM) with the OEM's incomplete information on the REM's unit cost. We apply the type-III model in Chapter 3 for formulation and derive the closed-form Bayesian Nash equilibrium. We use sensitivity analysis to investigate the effect of such incomplete information on both competitors' decision making.</p> <p>We summarize in Chapter 5 and provide a general direction for future research on game theoretic applications in RL.</p> / Doctor of Philosophy (PhD)

Positive biotic interactions such as mutualism, commensalism and facilitation are ubiquitous in nature, but historically have received considerably less research attention than negative interactions such as competition, predation and parasitism. The paucity of research on positive interactions is particularly evident in stream ecosystems and in vertebrate communities. Stream fishes clearly provide an ideal system for advancing research on positive interactions. Many minnows (Cyprinidae) of eastern North America engage in a potentially mutualistic reproductive interaction known as nest association, in which individuals of one species (nest associates) spawn in nests constructed by host species. In nest association, hosts provide unsilted gravel substrate for spawning nest associates, and increased parental care to associate broods. High associate: host egg ratios can create a dilution effect, reducing the probability that host eggs will be preyed upon by egg predators. Nest associative interactions are common, but are relatively understudied compared to other interactions among stream fishes. The goals of this study were to apply general ecological models to this novel system to (a) gain new insight into the mechanisms structuring nest associative stream fish communities, and (b) to use inference from stream fish communities to potentially expand and improve the general ecological models. These goals required completion of three objectives, including (1) examining the influence of abiotic and biotic contexts on reproductive behavior and fitness outcomes between a cyprinid host and associate, using the biological markets model to generate predictions; (2) examining the utility of the nest web framework (previously only used for cavity nesting vertebrate communities) and the stress gradient hypothesis (previously applied almost exclusively to plant communities) for predicting which associate species spawn on nests built by various nest building species, and the consequences of these choices, respectively; and (3) using two-species occupancy modeling to determine the relative influence of biotic interactions and habitat covariates on the co-occurrence of a host and two nest associates. To accomplish these goals, I conducted a large-scale experiment to manipulate presence of mutualists (Nocomis leptocephalus, host; Chrosomus oreas, associate), egg predators (biotic context) and habitat quality (abiotic context). I conducted behavioral nest observations and conducted repeated stream fish stream fish community surveys to collect demographic data. I constructed a nest web from observational data, and implemented structural equation modeling through an information-theoretic framework to identify nest web plausibility across a large spatial extent. I tested some predictions of the stress gradient hypothesis by regressing juveniles-per-nest and a metric of cyprinid community structure on a composite measure of physical stress (scaled gradients of catchment-scale agricultural land use and catchment area). I used two-species occupancy modeling to model co-occurrence of N. leptocephalus hosts and two associates, C. oreas and Clinostomus funduloides, and used an information-theoretic framework to compare hypotheses representing the importance of biotic interactions, habitat covariates or both at determining species co-occurrence. Results corroborated some (but not all) model predictions, and identified room for improvement in each of the general models. Nest associative spawning by C. oreas was not context dependent; C. oreas did not spawn in the absence of a reproductively active male N. leptocephalus at any treatment level. However, the net fitness outcome of host and associate species was mutualistic, and the interaction outcome switched from commensalistic to mutualistic with abiotic context. N. leptocephalus reproductive success was improved by C. oreas presence in less-silted habitats, but not in heavily-silted habitats. This is most likely because broods were subject to predation in both habitat types, but were also negatively affected by siltation in silted habitats. Accordingly, egg dilution by associates was not sufficient to support a mutualistic relationship in less favorable habitats. Results suggest that the biological markets model may be a useful tool for predicting fitness outcomes of nest associative mutualism, but may not be as useful for predicting the behavioral outcomes of obligate mutualisms. Future applications of the biological markets model should carefully consider species traits, specifically the degree to which trading behavior is obligate for participants. Future work with this model will yield more insight by considering highly facultative associates. Nest webs constructed from nest observational data suggested an interaction topology in which strong (nearly-obligate) associates relied most frequently on N. leptocephalus nests, and less frequently on nests constructed by Campostoma anomalum. Weak (facultative) associates were seldom associated with nests constructed by either species, and probably spawned before hosts began nesting activity. Structural equation models corroborated this topology throughout the New River basin, although some less-supported model evidence specified some nest association by weak associates. Juveniles-per-nest of strong associates responded positively to physical stress, while this metric for other cyprinid reproductive groups showed no relationship. Proportional representation of Nocomis and strong associates also increased predictably with physical stress. This study suggests that the nest web framework can be informative to systems outside the ones for which it was developed; future studies may be able to use this framework to better understand the role of habitat-modifying species in communities other than cavity nesting terrestrial vertebrates and nest associative stream fishes. This work extended the nest web framework by (a) modeling the outcomes of interactions instead of the interactions themselves, and (b) by using structural equation modeling to test nest web predictions with an information-theoretic framework. This study also suggests that the stress gradient hypothesis can be useful for understanding interaction dynamics in vertebrate communities; this represents the first direct evidence that this model can be used in vertebrate communities. Further, I demonstrate that the stress gradient hypothesis may be extended to predict community structure. However, more research in a diversity of systems will be needed to determine the extent to which this can be applied. This study provides some of the first evidence of large-scale positive co-occurrence patterns in vertebrates. However, the precise roles of habitat covariates and biotic interactions were species-specific. Occupancy results suggest that co-occurrence between N. leptocephalus and nest associate C. funduloides is driven only by reproductive behavioral interactions. Alternatively, evidence suggests that co-occurrence between N. leptocephalus and C. oreas is driven by both nest association and habitat covariates. That two-species occupancy modeling can be a useful tool for comparing difficult-to-test hypotheses involving biotic interactions at large spatial scales. This study represents the first quantitative, multi-scale treatment of positive interactions in stream ecosystems. This study demonstrates that applying general ecological models to stream fish communities can yield new insights about both the study system and the models themselves. While models of negative interactions, food webs and dispersal have been applied to stream fishes, we stand to gain much ground by also considering positive biotic interactions. In doing so, stream fish ecologists will also be able to contribute to the advancement of general ecology, and thus raise awareness for these understudied ecosystems and taxa. / Ph. D.

The accelerating pace of growth of aquaculture in sub-Saharan Africa has received much positive appraisal because of the potential of the industry to contribute to economic development and food security by providing jobs and animal protein. Adoption of best management practices (BMPs) holds the potential to ameliorate the related environmental impacts of aquaculture, such as in the amounts of nutrients and sediment that will enter natural water bodies from earthen pond effluents. The goals of this study were to characterize adoption of aquaculture BMPs on small-scale, pond-based farms in Ghana, and to assess selected economic, social, and environmental outcomes of BMP adoption. Two BMPs: 1) water reuse, and 2) commercial floating feeds, were investigated for adoption by pond-based fish farmers in Ghana. I conducted my study in Ghana using on-farm experiments involving intensive monitoring of water quality and growth of Nile tilapia (Oreochromis niloticus) over two production cycles. Additionally, I administered a baseline survey to 393 (and a follow-up survey to 160) fish farmers. I determined the best model for modelling farmed Nile tilapia growth with multi-model inference based on Akaike information criterion (AIC), the profitability of adopting BMPs with stochastic enterprise budgets and, social welfare impact with the Economic-Surplus model. I used a Markov model to predict the equilibrium rate of adoption of the two BMPs and determined the impact of BMP adoption on the reduction of pollutant loading with the Minimum-Data method of the Tradeoffs Analysis (TOA-MD). My results showed that the logistic model is a better alternative to the von Bertalanffy model for modelling the growth of Oreochromis niloticus under pond aquaculture conditions. There were no significant differences in fish weight between the water re-use BMP and the use of new water. Adoption of the commercial floating feed BMP resulted in a 100% increase in fish final weight and yield, and in higher profitability, compared to the sinking feed type. Probability of making a profit was highest (72%) in the scenario with commercial feed and self-financing. Net present values (NPV) of about US$ 11 million and US$ 375 million could be obtained from the adoption of commercial floating feed and Genetically-Improved Farm Tilapia (GIFT) strain, respectively, in Ghana. Hence, any innovation that has a significant impact on fish yield also will have a significant impact on mean NPV and social welfare. However, I identified a number of potential negative ecological and genetic impacts exist from introducing the GIFT strain into Africa from Asia. Although considered low-intensity production systems, nutrients and solids in study ponds were found to be higher than levels expected in intensive culture ponds by wide margins. Pond water quality was significantly higher with commercial floating feed. The water-reuse BMP also prevented pollutants from leaving ponds altogether for the number of cycles for which pond water was reused, especially if associated BMPs such as rainfall capture and avoidance of water exchange are observed. Significant reductions in the loading of all water quality variables (nitrogen, phosphorus, solids, and BOD5) could be achieved with the adoption of the recommended feed type in Ghana. Adoption of the water reuse BMP has the potential to cause pollution reductions of 200% - 3,200% above that from the floating feed BMP. The strongest influence on the combined adoption of these BMPs were from : farmer's awareness of the feed BMP, perceived necessity and relative profitability of the water reuse BMP, and farmer's years of experience. A combination of central media (workshops), demonstrations, and lateral diffusion was found to be the most effective channel for disseminating these BMPs. Maximum adoption rate of the feed BMPs was estimated to be 38% - 58%. Also, US$ 6,000/year and US$12,000/year need to be paid per 0.6 ha pond surface area to push adoption of the feed BMP to 50% and 70%, respectively. Hence, to ensure the successful adoption of aquaculture BMPs, I recommend that regular well-planned workshops be organized to create awareness and a conducive atmosphere to target farmers at multiple stages of the innovation decision process. Incentives and effective dissemination will encourage the adoption of these and other environmental BMPs. Feed costs need to be lowered in order to encourage the adoption of commercial floating feed in Ghana. Future analyses could quantify the differences in production costs between using the two water types, to reveal the possible higher relative profitability of pond water reuse over draining ponds after each production cycle. Also, African governments are advised to commission rigorous baseline and ecological risk analyses before adoption of the GIFT strain. Improvements in management practices and infrastructure could increase the yield and profitability of the local strains even if genetically-improved strains are not introduced. / Ph. D.

My thesis aims at contributing to classifying the Liar-like paradoxes (and related Truth-teller-like expressions) by clarifying distinctions and relationships between these expressions and arguments. Such a classification is worthwhile, firstly, because it makes some progress towards reducing a potential infinity of versions into a finite classification; secondly, because it identifies a number of new paradoxes, and thirdly and most significantly, because it corrects the historically misplaced distinction between semantic and set-theoretic paradoxes. I emphasize the third result because the distinction made by Peano [1906] and supported by Ramsey [1925] has been used to warrant different responses to the semantic and set-theoretic paradoxes. I find two types among the paradoxes of truth, satisfaction and membership, but the division is shifted from where it has historically been drawn. This new distinction is, I believe, more fundamental than the Peano-Ramsey distinction between semantic and set-theoretic paradoxes. The distinction I investigate is ultimately exemplified in a difference between the logical principles necessary to prove the Liar and those necessary to prove Grelling’s and Russell’s paradoxes. The difference relates to proofs of the inconsistency of naive truth and satisfaction; in the end, we will have two associated ways of proving each result. ¶ Another principled division is intuitively anticipated. I coin the term 'hypodox' (adj.: 'hypodoxical') for a generalization of Truth-tellers across paradoxes of truth, satisfaction, membership, reference, and where else it may find applicability. I make and investigate a conjecture about paradox and hypodox duality: that each paradox (at least those in the scope of the classification) has a dual hypodox.¶ In my investigation, I focus on paradoxes that might intuitively be thought to be relatives of the Liar paradox, including Grelling’s (which I present as a paradox of satisfaction) and, by analogy with Grelling’s paradox, Russell’s paradox. I extend these into truth-functional and some non-truth-functional variations, beginning with the Epimenides, Curry’s paradox, and similar variations. There are circular and infinite variations, which I relate via lists. In short, I focus on paradoxes of truth, satisfaction and some paradoxes of membership. ¶ Among the new paradoxes, three are notable in advance. The first is a non-truth functional variation on the Epimenides. This helps put the Epimenides on a par with Curry’s as a paradox in its own right and not just a lesser version of the Liar. I find the second paradox by working through truth-functional variants of the paradoxes. This new paradox, call it ‘the ESP’, can be either true or false, but can still be used to prove some other arbitrary statement. The third new paradox is another paradox of satisfaction, distinctly different from Grelling’s paradox. On this basis, I make and investigate the new distinction between two different types of paradox of satisfaction, and map one type back by direct analogy to the Liar, and the other by direct analogy to Russell's paradox.

Long-term diet patterns based on stable isotope analysis may be helpful to understand changes in food selection of black bears (Ursus americanus) over time and guide management programs to reduce human-bear conflicts. An enriched stable carbon isotope signature indicates an anthropogenic food source in the diet and an enriched nitrogen signature indicates a higher tropic level for a species. I examined longitudinal feeding patterns from 117 hair samples of black bears live captured in Great Smoky Mountains National Park during 1980–2001 using stable carbon and nitrogen isotope analysis from hair samples. I developed a set of a priori models to examine if sex, age class, year, weight class, total hard mast index, white oak index (Quercus spp.), red oak index (Quercus spp.), nuisance status and hog harvest (Sus scrofa) affected stable isotope signatures. I used model averaging and an estimator of the unconditional variance was used to account for model uncertainty. The δ[delta]13C signatures differed by weight class with above average weight, (ß[Beta] = 0.76‰; 95% CI = 0.28 to 1.23) and average weight (ß[Beta] = 0.42‰; CI = 0.06 to 0.78) showing enriched values compared to below average bears. Bears had enriched δ[delta]15N signatures in years with low white oak mast production (ß[beta] = -0.19, CI = -0.34 to -0.03) and depleted when white oak hard mast was abundant. Sub adult bears had enriched δ[delta]15N signatures compared to adult and older adult bears. Variation of nitrogen values was small during 1980–1991 ( = 2.57, SD = 0.28) but increased substantially during 1992–2000 ( = 2.29, SD = 0.71) when there was substantial variation in hard mast production. Bears in better physical condition appear more likely to access anthropogenic food sources. In years of low white oak acorn production, the larger bears and sub adult bears are more likely to turn to alternative food sources. The long term variation detected in this study is important in identifying which bears are potentially more likely to seek out the anthropogenic food sources when changes occur in availability of natural foods.

La grammaticalité d'une phrase est habituellement conçue comme une notion binaire : une phrase est soit grammaticale, soit agrammaticale. Cependant, bon nombre de travaux se penchent de plus en plus sur l'étude de degrés d'acceptabilité intermédiaires, auxquels le terme de gradience fait parfois référence. À ce jour, la majorité de ces travaux s'est concentrée sur l'étude de l'évaluation humaine de la gradience syntaxique. Cette étude explore la possibilité de construire un modèle robuste qui s'accorde avec ces jugements humains.<br>Nous suggérons d'élargir au langage mal formé les concepts de Gradience Intersective et de Gradience Subsective, proposés par Aarts pour la modélisation de jugements graduels. Selon ce nouveau modèle, le problème que soulève la gradience concerne la classification d'un énoncé dans une catégorie particulière, selon des critères basés sur les caractéristiques syntaxiques de l'énoncé. Nous nous attachons à étendre la notion de Gradience Intersective (GI) afin qu'elle concerne le choix de la meilleure solution parmi un ensemble de candidats, et celle de Gradience Subsective (GS) pour qu'elle concerne le calcul du degré de typicité de cette structure au sein de sa catégorie. La GI est alors modélisée à l'aide d'un critère d'optimalité, tandis que la GS est modélisée par le calcul d'un degré d'acceptabilité grammaticale. Quant aux caractéristiques syntaxiques requises pour permettre de classer un énoncé, notre étude de différents cadres de représentation pour la syntaxe du langage naturel montre qu'elles peuvent aisément être représentées dans un cadre de syntaxe modèle-théorique (Model-Theoretic Syntax). Nous optons pour l'utilisation des Grammaires de Propriétés (GP), qui offrent, précisément, la possibilité de modéliser la caractérisation d'un énoncé. Nous présentons ici une solution entièrement automatisée pour la modélisation de la gradience syntaxique, qui procède de la caractérisation d'une phrase bien ou mal formée, de la génération d'un arbre syntaxique optimal, et du calcul d'un degré d'acceptabilité grammaticale pour l'énoncé.<br>À travers le développement de ce nouveau modèle, la contribution de ce travail comporte trois volets.<br>Premièrement, nous spécifions un système logique pour les GP qui permet la révision de sa formalisation sous l'angle de la théorie des modèles. Il s'attache notamment à formaliser les mécanismes de satisfaction et de relâche de contraintes mis en oeuvre dans les GP, ainsi que la façon dont ils permettent la projection d'une catégorie lors du processus d'analyse. Ce nouveau système introduit la notion de satisfaction relâchée, et une formulation en logique du premier ordre permettant de raisonner au sujet d'un énoncé.<br>Deuxièmement, nous présentons notre implantation du processus d'analyse syntaxique relâchée à base de contraintes (Loose Satisfaction Chart Parsing, ou LSCP), dont nous prouvons qu'elle génère toujours une analyse syntaxique complète et optimale. Cette approche est basée sur une technique de programmation dynamique (dynamic programming), ainsi que sur les mécanismes décrits ci-dessus. Bien que d'une complexité élevée, cette solution algorithmique présente des performances suffisantes pour nous permettre d'expérimenter notre modèle de gradience.<br>Et troisièmement, après avoir postulé que la prédiction de jugements humains d'acceptabilité peut se baser sur des facteurs dérivés de la LSCP, nous présentons un modèle numérique pour l'estimation du degré d'acceptabilité grammaticale d'un énoncé. Nous mesurons une bonne corrélation de ces scores avec des jugements humains d'acceptabilité grammaticale. Qui plus est, notre modèle s'avère obtenir de meilleures performances que celles obtenues par un modèle préexistant que nous utilisons comme référence, et qui, quant à lui, a été expérimenté à l'aide d'analyses syntaxiques générées manuellement.

In this work, a multilayer security solution for digital communication systems is provided by considering the joint effects of physical-layer security channel codes with application-layer cryptography. We address two problems: first, the cryptanalysis of error-prone ciphertext; second, the design of a practical physical-layer security coding scheme. To our knowledge, the cryptographic attack model of the noisy-ciphertext attack is a novel concept. The more traditional assumption that the attacker has the ciphertext is generally assumed when performing cryptanalysis. However, with the ever-increasing amount of viable research in physical-layer security, it now becomes essential to perform the analysis when ciphertext is unreliable. We do so for the simple substitution cipher using an information-theoretic framework, and for stream ciphers by characterizing the success or failure of fast-correlation attacks when the ciphertext contains errors. We then present a practical coding scheme that can be used in conjunction with cryptography to ensure positive error rates in an eavesdropper's observed ciphertext, while guaranteeing error-free communications for legitimate receivers. Our codes are called stopping set codes, and provide a blanket of security that covers nearly all possible system configurations and channel parameters. The codes require a public authenticated feedback channel. The solutions to these two problems indicate the inherent strengthening of security that can be obtained by confusing an attacker about the ciphertext, and then give a practical method for providing the confusion. The aggregate result is a multilayer security solution for transmitting secret data that showcases security enhancements over standalone cryptography.

Traditional solutions to information security in communication systems act in the application layer and are oblivious to the effects in the physical layer. Physical-layer security methods, of which information-theoretic security is a special case, try to extract security from the random effects in the physical layer. In information-theoretic security, there are two asymptotic notions of secrecy---weak and strong secrecy This dissertation investigates the problem of information-theoretic strong secrecy on the binary erasure wiretap channel (BEWC) with a specific focus on designing practical codes. The codes designed in this work are based on analysis and techniques from error-correcting codes. In particular, the dual codes of certain low-density parity-check (LDPC) codes are shown to achieve strong secrecy in a coset coding scheme. First, we analyze the asymptotic block-error rate of short-cycle-free LDPC codes when they are transmitted over a binary erasure channel (BEC) and decoded using the belief propagation (BP) decoder. Under certain conditions, we show that the asymptotic block-error rate falls according to an inverse square law in block length, which is shown to be a sufficient condition for the dual codes to achieve strong secrecy. Next, we construct large-girth LDPC codes using algorithms from graph theory and show that the asymptotic bit-error rate of these codes follow a sub-exponential decay as the block length increases, which is a sufficient condition for strong secrecy. The secrecy rates achieved by the duals of large-girth LDPC codes are shown to be an improvement over that of the duals of short-cycle-free LDPC codes.

Made available in DSpace on 2016-08-17T14:53:18Z (GMT). No. of bitstreams: 1 Dissertacao Ewaldo.pdf: 1169300 bytes, checksum: fc5d4b9840bbafe39d03cd1221da615e (MD5) Previous issue date: 2012-02-10 / This work derives and evaluates a nonlinear method for Blind Source Extraction (BSE) in a Reproducing Kernel Hilbert Space (RKHS) framework. For extracting the desired signal from a mixture a priori information about the autocorrelation function of that signal translated in a linear transformation of the Gram matrix of the nonlinearly transformed data to the Hilbert space. Our method proved to be more robust than methods presented in the literature of BSE with respect to ambiguities in the available a priori information of the signal to be extracted. The approach here introduced can also be seen as a generalization of Kernel Principal Component Analysis to analyze autocorrelation matrices at specific time lags. Henceforth, the method here presented is a kernelization of Dependent Component Analysis, it will be called Kernel Dependent Component Analysis (KDCA). Also in this dissertation it will be show a Information-Theoretic Learning perspective of the analysis, this will study the transformations in the extracted signals probability density functions while linear operations calculated in the RKHS. / Esta dissertação deriva e avalia um novo método nãolinear para Extração Cega de Sinais através de operações algébricas em um Espaço de Hilbert Reproduzido por Kernel (RKHS, do inglês Reproducing Kernel Hilbert Space). O processo de extração de sinais desejados de misturas é realizado utilizando-se informação sobre a estrutura temporal deste sinal desejado. No presente trabalho, esta informação temporal será utilizada para realizar uma transformação linear na matriz de Gram das misturas transformadas para o espaço de Hilbert. Aqui, mostrarse- á também que o método proposto é mais robusto, com relação a ambigüidades sobre a informação temporal do sinal desejado, que aqueles previamente apresentados na literatura para realizar a mesma operação de extração. A abordagem estudada a seguir pode ser vista como uma generalização da Análise de Componentes Principais utilizando Kerneis para analisar matriz de autocorrelação dos dados para um atraso específico. Sendo também uma kernelização da Análise de Componentes Dependentes, o método aqui desenvolvido é denominado Análise de Componentes Dependentes utilizando Kerneis (KDCA, do inglês Kernel Dependent Component Analysis). Também será abordada nesta dissertação, a perspectiva da Aprendizagem de Máquina utilizando Teoria da Informação do novo método apresentado, mostrando assim, que transformações são realizadas na função densidade de probabilidade do sinal extraído enquanto que operação lineares são calculadas no RKHS.