No description available.

Dans cette thèse, nous considérons une famille des problèmes d’ordonnancement avec machine parallèle identique et contraintes de précédences. Ce champ de recherche fait l’objet de nombreuses études. Malgré tout, la complexité de ces problèmes varie selon de nombreux paramètres,notamment le type de graphe de précédence ou le critère retenu. De plus, il existe encore de nombreux problèmes ouverts. Nous étudions certains de ces problèmes dans cette thèse. Nous montrons notamment que le problème ouvert avec tâches de durée unitaires et graphe de précédence de type intree est NP-complet. Puis, nous prouvons que le problème avec graphe de précédence de type level order est NP-complet aussi. La preuve est ensuite étendue à des problèmes connexes. Par la suite, on améliore un algorithme exponentiel pour un problème spécifique qui est NP-complet. Enfin, nous proposons un modèle linéaire pour le problème avec contraintes de précédence quelconque, améliorant aussi les résultats de littérature. / The main problem studied in this thesis is that of parallel machine scheduling with precedence constraints. The complexity depends on the shape that the precedence graph takes and the objective function. We prove that one minimum-open problem of scheduling equal-processing-time jobs which subject to in-tree precedence constrains is NP complete while minimizing the total competition time.Then, we prove that the open problem of scheduling level-order precedence constrains is NP-complete too. We adapted the second proof to other scheduling problems as well.On the other hand, we improved an exponential algorithm designed for a specific NP-hard problem. At the end, we propose a linear programming model for the general scheduling problem with arbitrary precedence constraints and processing-time. We adapt the existing models which are originally designed for other scheduling problems to parallel scheduling problem and compare these models with ours.

In Re realism is the two-pronged view that, first, when this and that have the same color, this color and that color are identical. There is just one color, the universal. Second, on the view, this color exists just in case something has it. Say my cat has the same color as the dog I owned when I was a child. Since the dog existed before the cat, and precedence being irreflexive, it seems plausible to infer that the dog and the cat are distinct. Now take the colors. Since the colors are allegedly in re, and thus perhaps somehow elements of the cat and dog, it seems plausible to infer that the dog's color also preceded the cat's color. And therefore that the cat's color cannot be identical with the dog's. Finally, since the in re realist understands the sameness of properties in terms of identity, it follows that the cat's color cannot be the same as the dog's. The problem generalizes: What is the relationship between universals and time? Ignoring the temporality of that which constitutes time, to be temporal is to have a temporal "feature." These "features" are of three kinds: precedence, times, and being present, past, etc. The fundamental question in each case is whether universals have the feature. Do universals precede? Are they at times? Are they present? Time, I argue, is essentially the field in which things happen. To happen, I argue, is for one thing to do something. For one thing to do something is for the thing to exemplify a property. Such exemplifications of properties by objects I call "states of affairs." Only states of affairs precede, are at times, or are present. Universals, not being states of affairs, are not temporal. But, by the same argument which shows that running is not temporal it can be shown that Jack is not, even though Jack's running obviously is. So far I have defined what it is to be temporal; primitive temporality. But since Jack is a constituent of something temporal, he may be justly thought of as derivatively temporal.

In current age, various companies are using a centralized database system for dailybusiness transactions in different domains. Some critical issues have been observedrelated to the complexity, maintenance, performance and communication cost of datain centralized data repository for query processing, according to the demand of endusersfrom different locations. So, different enterprises are striving to implementefficient distributed database systems in their business environments for scalability.The distributed database architecture covers different factors such as transparentmanagement system, replication, fragmentation and allocation etc. This dissertationfocuses on database fragmentation and techniques which are useful for performingdatabase fragmentation. The objective of this research is to investigate efficient algorithm and technique fordatabase fragmentation in distributed environment. We proposed a customized ISUD(Insert, Select, Update, Delete) technique after comparative study of the best suitabletechniques, which is selected for implementation purpose. The functionality of thecustomized ISUD technique helps to get the precedence of the attribute of a relationhorizontally in database from various sites or location. The practical objective of this dissertation is to design the architecture and develop,implement customized ISUD (Insert, Select, Update, Delete) user interface, and to testthe selected algorithm or technique by using the interface. We used C#.Net as adevelopment tool. This user interface accepts ISUD frequency as an input andproduces ALP (attribute location precedence) values as output. We have incorporateddesign science research (DSR) method for customized ISUD technique development.This customized ISUD technique can be considered as a foundation to implementhorizontal database fragmentation in distributed environment, so that the databaseadministrator can take a proper decision for allocating the fragmented data to varioussites at initial state of distributed database design.

国立情報学研究所で電子化したコンテンツを使用している。

Thesis (J.C.D.)--Catholic University of America, 1932. / "Nota biográfica." "Bibliografía": p. 135-137.

Mate competition is a prominent component of sexual selection theory. Typically, males attempt to mate with the most females possible and females attempt to mate with the highest quality males possible. In the polygynous female-defense mating system of Anolis carolinensis, males compete directly for females through territorial behavior. Inter-male competition is intense due to an average polygyny ratio of 1 male to 3 females despite a 1:1 adult sex ratio. Through high levels of territorial behavior (e.g., 100 displays/h, 27 m patrol distances/h, 70% of day in defense-related activities), males attempt to exclude other males from resident females who, in turn, both store sperm and ovulate a single-egg clutch at weekly intervals over a 4-month breeding season. Paternity of hatchlings in 16 naturally occurring breeding groups was analyzed to determine the extent to which the territorial resident male was able to prevent other males from fathering offspring of his resident females. Lizards residing in or neighboring a resident male's territory were collected and RAPD-PCR was used to determine the paternity of hatchlings. Of the 48 hatchlings from 26 females, resident territorial males fathered 52% of hatchlings; 15% were fathered by a male whose territory bordered that of the resident male and 21% were fathered by a smaller male living covertly within the resident male's territory. Paternity for the remaining 12% of hatchlings belonged to an unsampled male. Given that females mated with multiple males, laboratory-based controlled matings were conducted where females were sequentially paired with two males and RAPD-PCR was used to analyze which of the two males fathered the subsequent hatchlings to determine the mechanism of sperm precedence. Regardless of mating order, only one male of the pair fertilized the eggs. Male A. carolinensis have reproductive strategies present in addition to defending resident females and female A. carolinensis have options in addition to simply mating with the resident male. While sperm precedence is present in this species, it is not based on mating order, but may involve both the number of sperm deposited in the female's tract as well as the quality of those sperm. / Ph. D.

In this thesis, we propose a family of concurrency control protocols for high data contention database environments. The first one is called the Prudent-Precedence Concurrency Control (PPCC) protocol. It is prudently more aggressive in permitting more serializable schedules than two-phase locking and maintains a restricted precedence among conflicting transactions and commits the transactions according to the serialization order established in the executions. The second one is a family of hierarchical concurrency control protocols called the Hierarchical Precedence Concurrency Control (HPCC) protocols. It maintains cycle-free precedence hierarchies for conflicting transactions. Conflicting operations are allowed to proceed only if the hierarchical orderings of precedence is not violated. Transactions also commit based on the serialization order established during the executions. Detailed simulation models have been implemented for all these protocols and extensive experiments have been conducted to evaluate the performance of the proposed approaches. The results demonstrate that the proposed algorithms outperform the two-phase locking and optimistic concurrency control over a wide range of system workloads.

<p>Rad,Edukativni i op&scaron;ti model kritičnih protoka materijala PD - Precedence Diagramming Structure, predstavlja integralni deo teorijskih i eksperimentalnih istraživanja vezanih za razvoj stohastičkih mreža PD-modela iz operacionih istraživanja, kao i modela za odgovarajuću edukaciju kadrova iz oblasti mrežnog upravljanja tokovima materijala u ma&scaron;inogradnji. U tom smislu u radu su razvijeni i re&scaron;avani, putem analitičkih i numeričkih metoda, sledeći modeli:</p><ul><li>Mrežni PD-model sa jednim kritičnim tokom.</li><li>Mrežni PD-model sa isključivo autonomnim (paralelnim) kritičnim tokovima.</li><li>Op&scaron;ti model sa varijantom kritičnih tokova autonomno-unijatnog tipa.</li><li>Edukativni model za efikasno re&scaron;avanje postavljenih problema upravljanjatokovima materijala u pojedinačnoj proizvodnji.</li><li>Modeli za vrednovanje metoda i postupaka iz ovih oblasti operacionih istraživanja, kao i samog edukativnog modela, na bazi razvijenih kvantitativnih i kvalitativnih kriterijuma.</li></ul><p>Ovim istraživanjem stvorene su određene podloge za dalji rad u predmetnom području, posebno sa stanovi&scaron;ta neinvesticionog povećanja efekata materijalne proizvodnje, putem primene razvijenih opercionih modela i edukativnih metoda za njihovo re&scaron;avanje.</p> / <p>The paper &quot;THE EDUCATIVE AND GENERAL MODEL OF THE CRITICAL MATERIAL FLOWS, PD-PRECEDENCE DIAGRAMMING STRUCTURE&quot; makes the integral part bof the theoretical and experimental researches connected for the development of the stochastic network PD - models of the opertions research, as well the models for the appropriote personal cadres edu cation in the field of network control of materijal flows in production of machinery. In the paper are also developed and solved, by analytic and numeric methods, the next models:</p><p>* The network PD-model with one critical flow.</p><p>* The network PD-model with excusively autonomus (parallel) critical flows.</p><p>* The general model with the variant of the critical flows of the autonomus -uniate type.</p><p>* The educative model for the effective solving of the stated problems of the material flows control in isolated production.</p><p>* The models and methods valuation of these fields of the operations researches, as well the educative model itself on the basic of the developed quantative and qualitative criteria.</p><p>With these researches are created the defined bases for further work in the subject field, particularly from the attitude of the noninvestment increasing of the effects of the materijal production, using the developed operations models and the educative models for their solving.</p>

Unmanned aerial vehicles (UAVs) are becoming increasingly popular for surveillance in civil and military applications. Vehicles built for this purpose vary in their sensing capabilities, speed and maneuverability. It is therefore natural to assume that a team of UAVs given the mission of visiting a set of targets would include vehicles with differing capabilities. This paper addresses the problem of assigning each vehicle a sequence of targets to visit such that the mission is completed with the least "cost" possible given that the team of vehicles is heterogeneous. In order to simplify the problem the capabilities of each vehicle are modeled as cost to travel from one target to another. In other words, if a vehicle is particularly suited to visit a certain target, the cost for that vehicle to visit that target is low compared to the other vehicles in the team. After applying this simplification, the problem can be posed as an instance of the combinatorial problem called the Heterogeneous Travelling Salesman Problem (HTSP). This paper presents a transformation of a Heterogenous, Multiple Depot, Multiple Traveling Salesman Problem (HMDMTSP) into a single, Asymmetric, Traveling Salesman Problem (ATSP). As a result, algorithms available for the single salesman problem can be used to solve the HMDMTSP. To show the effectiveness of the transformation, the well known Lin-Kernighan-Helsgaun heuristic was applied to the transformed ATSP. Computational results show that good quality solutions can be obtained for the HMDMTSP relatively fast. Additional complications to the sequencing problem come in the form of precedence constraints which prescribe a partial order in which nodes must be visited. In this context the sequencing problem was studied seperately using the Linear Program (LP) relaxation of a Mixed Integer Linear Program (MILP) formulation of the combinatorial problem known as the "Precedence Constrained Asymmetric Travelling Salesman Problem" (PCATSP).