AN ABSTRACT OF THE THESIS OF SOROOSH SOHANGIR, for the MASTER OF SCIENCE degree in COMPUTER SCIENCE, presented on 9 th November 2011, at Southern Illinois University Carbondale. TITLE: OPTIMIZED FEATURE SELECTION USING NEUROEVOLUTION OF AUGMENTING TOPOLOGIES (NEAT) MAJOR PROFESSOR: Dr. Shahram Rahimi Feature selection using the NeuroEvolution of Augmenting Topologies (NEAT) is a new approach. In this thesis an investigation had been carried out for implementation based on optimization of the network topology and protecting innovation through the speciation which is similar to what happens in nature. The NEAT is implemented through the JNEAT package and Utans method for feature selection is deployed. The performance of this novel method is compared with feature selection using Multilayer Perceptron (MLP) where Belue, Tekto, and Utans feature selection methods is adopted. According to unveiled data from this thesis the number of species, the training, accuracy and number of hidden neurons are notably improved as compared with conventional networks. For instance the time is reduced by factor of three.

Multi-core architectures are the future for high-performance computing and are omnipresent these days; what was a vision some twenty years back is now a reality with most personal computers/laptops now running on multi-cores making them ubiquitous in today's world. However, as the number of cores continue scaling with time, there will be serious throughput and performance issues with relation to the network topologies used in connecting the cores. Among possible network topologies under consideration in modern multi-core systems, the `Mesh' topology is widely used. In terms of performance, the `Point to Point topology' would outperform all other topologies such as Crossbar, Mesh and Torus. The `Point to Point' topology does include additional expenses with respect to more links needed to connect each core to every other core in the network. Its expensive implementation cost is the reason it is not preferred in the industry for general use systems. But, for research purposes it serves as the best network topology alternative to the `Mesh' for higher speed in computer systems. However, the characteristics of the tasks executing on the cores will also have a significant impact on topology performance. So, with the scaling of multi-cores from 10 to 1000 cores per chip and more, selection of the right network topology is of importance. Another interesting factor to consider is the effect of the cache on these multi-core systems with respect to each of these topologies. Cache coherency is and will be a major cause for throughput decrease as cores scale. In our work, we are using the Modified-Exclusive-Shared-Invalid (MESI) Cache Coherency protocol for all the above mentioned network topologies considered. In this thesis, we investigate the effect of varying cache parameters such as the sizes of L1 Instruction cache, L1 Data cache and L2 cache and their respective associativities on each network topology. Various combinations of all these four parameters were considered as we ran experiments. We use the gem5 Computer Architecture Simulator for running our experiments with 4 core models. For benchmark purposes, we use the SPLASH-2 set of {\it `High Performance Computing'} benchmarks. A benchmark is assigned to each core. We also observe the effects of running benchmarks with similar characteristics on all cores versus comparing them with a set of different benchmarks while keeping all other parameters constant. Through our results, we attempt to give researchers and the industry at large a better view of the advantages and disadvantages along with the relationship between multi-cores, the cache and network topologies for multi-core systems.

This paper examines several questions regarding Banach spaces, completeness and compactness of Banach spaces, dual spaces and weak and weak* topologies. Examples of completeness and isometries are given using the c₀ and 𝓁ᴰ spaces. The Hahn-Banach extension theorem is presented, along with some applications. General theory about finite and infinite dimensional normed linear spaces is the bulk of the second chapter. A proof of the uniform boundedness principle is also given. Chapter three talks in detail about dual spaces and weak and weak* topologies. An embedding proof and proofs involving weak and weak compactness are also given. The Cauchy-Bunyakowski-Schwarz inequality and Alaoglu's theorem are also proven.

No description available.

Living a sustainable lifestyle while facing increasing population and decreasing natural resources has become one of humanity’s largest challenges. Locating fossil fuels is becoming more difficult while the demand for them to power our societies is ever increasing. Instead of finding more efficient methods of extracting fossil fuels, developing technologies that create renewable substitutes for fossil fuels is now the strategy. Algae biofuel matches fossil fuel performance while also meeting the criteria for renewable energy. The focus now shifts to finding methods for commercially producing algae biofuel. Therefore, the objective of this thesis is to develop a system that provides the flexibility in finding the optimum operating conditions for lysing algae. Lysing is the process of disrupting the cell membrane in order to isolate the cellular components necessary to produce biofuel. The proposed system consists of cascaded power converters that provide a pulse output voltage in order to create a pulsed electric field (PEF) to lyse algae. The proposed system is unique from any known PEF systems because it provides the ability to independently adjust peak voltage, pulse width and frequency of the output voltage. This in turn provides great flexibility in determining optimum pulse voltage at various operating conditions for lysing algae. The system was tested on its ability to control the required variables while maintaining independence from the other variables. The new network was also designed and tested on how well it regulated the specific output waveform under the effects of different load currents as well as variations in the input voltage.

The generalization ability of artificial neural networks (ANN) is highly dependent on their architectures and can be critical to solving a given problem. The current best practice uses fixed architectures determined via a trial-and-error approach. This process can be both computationally and temporally cumbersome and does not guarantee that an optimal topology will even be found. Replacing the user’s role in designing topologies with methods that enable a system to manage its own growth can endow systems with adaptable learning. Constructive algorithms offer the possibility of compact architectures as an alternative to the trial-and-error approach. This class of algorithms grows a network’s topology by incrementally adding units during learning to match task complexity. However, the decision of when to add new units in constructive algorithms heavily depends on user-defined a priori hyperparameters, which can be task-specific. Contrary to having a user fine-tune hyperparameters that govern growth, the intrinsic population dynamics of an ANN could be used to self-govern the growing process. Theoretically, an ANN or each layer comprising the network can be viewed as a set of populations. From this perspective, a hidden layer can be considered the environment in which hidden units exist. In this work, we propose a novel, more self-governed growing algorithm inspired by population dynamics for determining near-optimal topologies of feedforward ANNs. This allows the inclusion of a carrying capacity, the maximum population of hidden units that can be sustained in a hidden layer. Including this constraint in combination with population dynamics provides a built-in mechanism for a dynamic growth rate. The proposed approach is used in parallel with direct performance feedback from the network to modulate the growth rate of the hidden layer, allowing the network to converge to smaller topologies based on the task's demands. More self-governed approaches reduce the number of finely-tuned hyperparameters required to decide when to grow and put more control of the network’s structure and representational capacities in the algorithms themselves, facilitating the emergence of inherent intelligent behaviour. Chapter one introduces a dynamic, more self-governed growing algorithm inspired by population dynamics. Results show that compared to using fixed rules for determining hidden layer sizes; dynamic growth leads to smaller topologies than predicted while still being capable of solving the task. In chapter two, we investigate the algorithm's inherent properties to validate the more self-governed aspect. The results depict that the model’s hyperparameters require less fine-tuning by the user and adhere more toward self-governance. Finally, in chapter three, we investigate the effects of growing hidden layers individually in a sequential fashion or simultaneously in a parallel fashion multilayer context. A modified version of the growing algorithm capable of growing parallel is proposed. Growing hidden layers in parallel resulted in comparable or higher performances than sequential approaches. The growing algorithm presented here offers more self-governed growth, which provides an effective general solution automatically tailored to the task.

No description available.

De multiples questions se posent sur les liens sociaux. Parmi elles, le présent travail s’intéresse plus spécifiquement à leur orchestration à l’échelle individuelle : comment s’agencent ou sont agencés les liens sociaux qu’entretient au quotidien un individu ? En tant que reflets de ces liens, les mobilités tant réelles (déplacements) que virtuelles (pratiques des Technologies de l’Information et de la Communication) permettent de mettre en évidence les façons dont les liens privés et professionnels s’articulent, se positionnent les uns par rapport aux autres, entre superposition et séparation. Outre l’idéal hypermoderne du fluide ou superposition totale, d’autres topologies aux moindres degrés d’ouverture existent : la région et le réseau. L’existence de ces trois figures topologiques montre que l’agencement des sphères et des liens sociaux vu à travers les pratiques de mobilités est une activité complexe. Cette activité est le produit même des liens sociaux individuels. Ces derniers jouent dans leur dimension verticale (les normes) et dans leur dimension horizontale (les relations interpersonnelles). Toutefois, en tant qu’élément central de ces liens, l’individu joue un rôle important dans le travail d’orchestration du quotidien / There are many questions which are held about social ties. Here we aim at focusing more especially on the way these ties are organised at the individual scale: how are social ties arranged for individuals in everyday life? As reflection of social ties, real (i.e, moves) and virtuals (i.e. ICT’s uses) mobilities provides means to understand how private and professional ties are balanced between superposition and separation. In fact, moreover the hypermodern ideal of fluid (or total superposition), other topologies which have various degrees of opening exist: the network and the region. These three co-existent topologies show that the agency of spheres and social ties among mobile practices is a complex task. It is the product of social ties themselves. They are effective in their vertical (norms) and horizontal (relationships) dimensions. Moreover, as the core element of the described ties, the individual himself is an important part of this daily work of orchestration

In a domain with diverse multi-disciplinary views of what a Wireless Sensor Network (WSN) is, tracking progress and developing efficient WSNs is inherently a complex process. The main motivation of this work is advancing state-of-the-art WSNs by adaptively utilizing their components, and enlisting the utility of resources in network vicinity. As WSNs increase in density and expand in scale, we continue to witness an increase in overlapped deployments that serve independent applications. In most scenarios, new networks are deployed for new applications without considering previous or neighboring WSNs. This thesis presents the resource reuse (RR-WSN) paradigm. Adopting a generic framework for resource utilization, we achieve synergy between heterogeneous sensing systems. We abstract the view of a WSN in terms of functional capabilities, and offer a component-based view to boost sensor node (SN) potential and contribution to WSN operation. Thus SNs provide resources. On the other hand, we formally derive a set of functional requirements per application. The design and deployment of WSNs thus converges to an optimal assignment of functional requirements to resources. Two mainstream designs of WSNs are addressed in this thesis. The first involves WSNs with static deployments of nodes, whereby multiple applications run on networks in a given vicinity, yet the resources and applications share an owner (e.g., on a University Campus). We then present a Binary Integer Programming formulation to find the optimal assignment of resources to these functional requirements, while minimizing the energy impact of running each functional request. We further extend our scope to include WSNs that depend on transient nodes, such as smartphones, in a dynamic (DRR-WSN) paradigm, which could contribute significantly to the resource pool. Intuitively, multiple-owners are involved as resource providers and require different applications. Thus, we address the valuation of resources as they are shared across network owners. We finally present a maximal matching problem of finding the lowest cost for running each application, based on the available resource pool in the vicinity required. Extensive performance evaluation depicts the impact of RR-WSN design on WSN operation and longevity in various scenarios. / Thesis (Ph.D, Computing) -- Queen's University, 2013-08-27 04:44:14.556

Seja (X; t) um espaço topológico e seja F a família de todos os subconjuntos de X que satisfazem uma propriedade topológica dada P (invariante por homeomorfismos). Se acrescentarmos abertos novos à topologia e se F\' é a família de todos os subconjuntos do novo espaço que satisfazem a propriedade P, podemos ter que F ≠ F\'. Se isto sempre acontece, dizemos que o espaço (X; t) é maximal com respeito à família F. Neste trabalho estudaremos os espaços topológicos maximais com respeito a algumas famílias de subconjuntos: discretos, compactos, densos, conexos e das sequências convergentes. / Let (X; t) be a topological space and let F be the family of all subsets of X that satisfy a given topological property P (invariant under homeomorphisms). If we add new open sets to the topology and if F\' is the family of all subsets of the new space which satisfy the property P, we can have F ≠ F\'. If this is always the case, we say that (X; t) is maximal with respect to the family F. We show here some characterizations of maximal spaces with respect to the family of some of its subsets: compacts, dense, discrete and convergent sequences.