Recently-developed technologies for monitoring activity in populations of neurons make it possible for the first time, in principle, to ask many basic questions in neuroscience. However, computational tools for analyzing newly available data need to be developed. The goal of this thesis is to contribute to this effort by focusing on two specific problems. First, we used a point-process regression framework to provide a methodology for statistical assessment of the link between neural spike synchrony and network-wide oscillations. In simulations, we showed that our method can recover ground-truth relationships, and in two types of spike train data we illustrated the kinds of results the method can produce. The approach improves on methods in the literature and may be adapted to many different experimental settings. Second, we considered the problem of source extraction in calcium imaging data, i.e., the detection of neurons within a field of view and the extraction of each neuron’s activity. The data we mainly focus on are recorded with a microendoscope, which has the unique advantage of imaging deep brain regions in freely behaving animals. These data suffer from high levels of background fluorescence, as well as the potential for overlapping neuronal signals. Based on the existing constrained nonnegative matrix factorization (CNMF) framework, we developed an efficient method to process microendoscopic data. Our method utilizes a novel algorithm to initialize the spatial shapes and temporal activity of the neurons from the raw video data independently from the strong fluctuating background. This step ensures the efficiency and accuracy of solving a nonconvex CNMF problem. Our method also models the complicated background by including its low-spatial frequency structure and the locally-low-rank feature to avoid absorbing cellular signals into the background term. We developed a tractable solution to estimate the background activity using this new model. After subtracting the approximated background, we followed the CNMF framework to demix neural signals and recover denoised and deconvolved temporal activity. We optimized several algorithms in solving the CNMF problems to get accurate results. In practice, our method outperforms all existing methods and has been adopted by many experimental labs.

This thesis focuses on international intervention and statebuilding in post-2001 Afghanistan. It offers an alternative lens, a network lens, to understand the complexity of internationally sponsored state re-building and transformation. It therefore analyses how political power is assembled and flows through political networks in statebuilding, with an eye to the hitherto ignored endogenous political networks. The empirical chapters investigate the role and power dynamics of Afghan political network in re-assembling and transforming the post-2001 state once a political settlement is reached; how everyday political network practices shape the nature of statehood and governance; and subsequently how these power dynamics and practices contribute towards political order/violence and stability/instability. This thesis challenges the dominant wisdom that peacebuilding is a process of democratisation or institutionalisation, showing how intervention has unintentionally produced the democratic façade of a state, underpinning by informal power structures of Afghan politics. The post-2001 intervention has fashioned a ‘network state’ where the state and political networks have become indistinguishable from one another: the empowered network masquerade as the state. This study suggests that a new political order is emerging in post-2001 Afghanistan where political stability is a function of patron-client relations, opportunistic practices of bargaining and expropriation of public resources for political network gain as well as the instrumentalisation of identities. In light of this analysis, it concludes with the implications of the research findings for the future of Afghanistan. It posits that a successful international military exit from Afghanistan and post-2014 state survival may depend primarily on the political stability of the empowered political networks. This research is based on extensive fieldwork, including participatory observation and interviews (more than 130 interviews) with key informants over 16 months in Afghanistan.

abstract: Transfer learning refers to statistical machine learning methods that integrate the knowledge of one domain (source domain) and the data of another domain (target domain) in an appropriate way, in order to develop a model for the target domain that is better than a model using the data of the target domain alone. Transfer learning emerged because classic machine learning, when used to model different domains, has to take on one of two mechanical approaches. That is, it will either assume the data distributions of the different domains to be the same and thereby developing one model that fits all, or develop one model for each domain independently. Transfer learning, on the other hand, aims to mitigate the limitations of the two approaches by accounting for both the similarity and specificity of related domains. The objective of my dissertation research is to develop new transfer learning methods and demonstrate the utility of the methods in real-world applications. Specifically, in my methodological development, I focus on two different transfer learning scenarios: spatial transfer learning across different domains and temporal transfer learning along time in the same domain. Furthermore, I apply the proposed spatial transfer learning approach to modeling of degenerate biological systems.Degeneracy is a well-known characteristic, widely-existing in many biological systems, and contributes to the heterogeneity, complexity, and robustness of biological systems. In particular, I study the application of one degenerate biological system which is to use transcription factor (TF) binding sites to predict gene expression across multiple cell lines. Also, I apply the proposed temporal transfer learning approach to change detection of dynamic network data. Change detection is a classic research area in Statistical Process Control (SPC), but change detection in network data has been limited studied. I integrate the temporal transfer learning method called the Network State Space Model (NSSM) and SPC and formulate the problem of change detection from dynamic networks into a covariance monitoring problem. I demonstrate the performance of the NSSM in change detection of dynamic social networks. / Dissertation/Thesis / Doctoral Dissertation Industrial Engineering 2015

Broadly this thesis looks into network and time-series mining problems pertaining to dynamics over networks in various domains. Which locations and staff should we monitor in order to detect C. Difficile outbreaks in hospitals? How do we predict the peak intensity of the influenza incidence in an interpretable fashion? How do we infer the states of all nodes in a critical infrastructure network where failures have occurred? Leveraging domain-based information should make it is possible to answer these questions. However, several new challenges arise, such as (a) presence of more complex dynamics. The dynamics over networks that we consider are complex. For example, C. Difficile spreads via both people-to-people and surface-to-people interactions and correlations between failures in critical infrastructures go beyond the network structure and depend on the geography as well. Traditional approaches either rely on models like Susceptible Infectious (SI) and Independent Cascade (IC) which are too restrictive because they focus only on single pathways or do not incorporate the model at all, resulting in sub-optimality. (b) data sparsity. Additionally, the data sparsity still persists in this space. Specifically, it is difficult to collect the exact state of each node in the network as it is high-dimensional and difficult to directly sample from. (c) mismatch between data and process. In many situations, the underlying dynamical process is unknown or depends on a mixture of several models. In such cases, there is a mismatch between the data collected and the model representing the dynamics. For example, the weighted influenza like illness (wILI) count released by the CDC, which is meant to represent the raw fraction of total population infected by influenza, actually depends on multiple factors like the number of health-care providers reporting the number and public tendency to seek medical advice. In such cases, methods which generalize well to unobserved (or unknown) models are required. Current approaches often fail in tackling these challenges as they either rely on restrictive models, require large volume of data, and/or work only for predefined models. In this thesis, we propose to leverage domain-based frameworks, which include novel models and analysis techniques, and domain-based low dimensional representation learning to tackle the challenges mentioned above for networks and time-series mining tasks. By developing novel frameworks, we can capture the complex dynamics accurately and analyze them more efficiently. For example, to detect C. Difficile outbreaks in a hospital setting, we use a two-mode disease model to capture multiple pathways of outbreaks and discrete lattice-based optimization framework. Similarly, we propose an information theoretic framework which includes geographically correlated failures in critical infrastructure networks to infer the status of the network components. Moreover, as we use more realistic frameworks to accurately capture and analyze the mechanistic processes themselves, our approaches are effective even with sparse data. At the same time, learning low-dimensional domain-aware embeddings capture domain specific properties (like incidence-based similarity between historical influenza seasons) more efficiently from sparse data, which is useful for subsequent tasks. Similarly, since the domain-aware embeddings capture the model information directly from the data without any modeling assumptions, they generalize better to new models. Our domain-aware frameworks and embeddings enable many applications in critical domains. For example, our domain-aware frameworks for C. Difficile allows different monitoring rates for people and locations, thus detecting more than 95% of outbreaks. Similarly, our framework for product recommendation in e-commerce for queries with sparse engagement data resulted in a 34% improvement over the current Walmart.com search engine. Similarly, our novel framework leads to a near optimal algorithms, with additive approximation guarantee, for inferring network states given a partial observation of the failures in networks. Additionally, by exploiting domain-aware embeddings, we outperform non-trivial competitors by up to 40% for influenza forecasting. Similarly, domain-aware representations of subgraphs helped us outperform non-trivial baselines by up to 68% in the graph classification task. We believe our techniques will be useful for variety of other applications in many areas like social networks, urban computing, and so on. / Doctor of Philosophy / Which locations and staff should we monitor to detect pathogen outbreaks in hospitals? How do we predict the peak intensity of the influenza incidence? How do we infer the failures in water distribution networks? These are some of the questions on dynamics over networks discussed in this thesis. Here, we leverage the domain knowledge to answer these questions. Specifically, we propose (a) novel optimization frameworks where we exploit domain knowledge for tractable formulations and near-optimal algorithms, and (b) low dimensional representation learning where we design novel neural architectures inspired by domain knowledge. Our frameworks capture the complex dynamics accurately and help analyze them more efficiently. At the same time, our low-dimensional embeddings capture domain specific properties more efficiently from sparse data, which is useful for subsequent tasks. Similarly, our domain-aware embeddings are inferred directly from the data without any modeling assumptions, hence they generalize better. The frameworks and embeddings we develop enable many applications in several domains. For example, our domain-aware framework for outbreak detection in hospitals has more than 95% accuracy. Similarly, our framework for product recommendation in e-commerce for queries with sparse data resulted in a 34% improvement over state-of-the-art e-commerce search engine. Additionally, our approach outperforms non-trivial competitors by up to 40% in influenza forecasting.

The default mode network (DMN) is a highly cited neural network whose functional roles are not well understood. Until recently, event related fMRI experiments used to study the DMN could only be conducted in an open-loop format. The purpose of this study was to demonstrate the potential statistical advantages of real-time fMRI studies to conduct closed-loop experiments to directly test putative DMN functions. Using both fMRI simulations and large archival datasets, we demonstrate that open-loop designs are less statistically powerful than closed-loop experiments that can trigger stimuli at controlled levels of brain activity. When simulating event scheduling on resting state data, DMN levels were normally distributed, but the event timing proved to be ineffective in capturing the highest and lowest DMN values on average across subjects. Statistical differences in DMN levels collected by the Human Connectome Project-Aging (HCP-A) during a Go/NoGo task were also reported, along with the network's distributional effects across subjects. When examining DMN levels in 136 subjects more prone to commission errors the mean DMN levels were reported to be higher during and prior to incorrect NoGo responses. Exploring DMN levels in these same individuals reacting to a Go task also revealed differing measurement patterns when compared to all 711 subjects in the study. Additionally, the distribution of total DMN levels across all participants, as well as during a Go or NoGo trial, showed a shift in the mean towards deactivation. Furthermore, the peak at this location was greater and revealed that increased sampling occurred at the mean and under sampling at the tails. Overall, the cumulative findings in this study were successful in providing statistical arguments to support propositions for more powerful closed-loop experimentation in fMRI. / Master of Science / Activity in a neural network is observed through the use of functional MRI (fMRI) by tracking higher levels of oxygenated blood to that region when active and lower quantities when inactive. Neural networks vary in their responsibilities, thus fMRI tasks are designed to trigger a response based on the functional role of the network. This can be exemplified by studying the blood flow to default mode network (DMN), a network responsible for mind wandering, during a task that requires focus. Researchers can then correlate moments of high activity, which indicates a greater degree of mind wandering, or low activity to a correct or incorrect response to the task. Unfortunately, the timing in which a task is presented to the participant is predetermined prior to the subject entering the MRI making it difficult to capture a correct or incorrect response at the precise moment of activation or deactivation. This concept is known as open-loop and often collects data at moments of neutral activity, neither high nor low. In contrast, a closed-loop design allows a researcher to monitor the DMN's activation levels in real time and present the task at a desired time. This provides more useful data to the experimenter as all recorded responses to the task correlate with exact moments of high and low activation. This makes claims about the neural network's role statistically more powerful as there is a greater quantity of data at these moments rather than during a neutral activation state. The purpose of this thesis is to provide statistical arguments that support propositions for more powerful closed-loop experimentation in fMRI.

Made available in DSpace on 2016-04-25T20:21:11Z (GMT). No. of bitstreams: 1 Wilken David Sanches.pdf: 3802834 bytes, checksum: 25eb0ebe2f508ee3a5f9d8a60f82048b (MD5) Previous issue date: 2014-04-16 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / This study aims to demonstrate the presence of the Free Software Movement ideas within the Brazilian state, by presenting how the ideology of the free sharing knowledge, defended by its activists, has been influencing public policies of digital inclusion, cultural production access, and the debate about the modernization of the Brazilian Copyright Law. Also, we aim to show how the dispute on the access to knowledge, in the international context, has been deviating, at least, temporarily, the Copyright Law from the centre of the public debate, putting in its place, the legislative procedures that regulate the use and the management of the internet / Este estudo busca evidenciar a trajetória do Movimento do Software Livre dentro do Estado brasileiro, mostrando de que maneira os ideais de livre circulação do conhecimento, apregoados por estes ativistas, têm influenciado as políticas públicas de inclusão digital, de acesso à produção cultural e o debate sobre a modernização da lei de direitos autorais no país. Além disso, demonstrar como as disputas pelo acesso ao conhecimento, aliadas à conjuntura internacional, retiram, momentaneamente, a lei de direitos autorais do centro da disputa, colocando em seu lugar os processos legislativos que regulamentam do uso e a gestão da rede mundial de computadores

Made available in DSpace on 2016-04-26T14:54:49Z (GMT). No. of bitstreams: 1 Wilken David Sanches.pdf: 3802834 bytes, checksum: 25eb0ebe2f508ee3a5f9d8a60f82048b (MD5) Previous issue date: 2014-04-16 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / This study aims to demonstrate the presence of the Free Software Movement ideas within the Brazilian state, by presenting how the ideology of the free sharing knowledge, defended by its activists, has been influencing public policies of digital inclusion, cultural production access, and the debate about the modernization of the Brazilian Copyright Law. Also, we aim to show how the dispute on the access to knowledge, in the international context, has been deviating, at least, temporarily, the Copyright Law from the centre of the public debate, putting in its place, the legislative procedures that regulate the use and the management of the internet / Este estudo busca evidenciar a trajetória do Movimento do Software Livre dentro do Estado brasileiro, mostrando de que maneira os ideais de livre circulação do conhecimento, apregoados por estes ativistas, têm influenciado as políticas públicas de inclusão digital, de acesso à produção cultural e o debate sobre a modernização da lei de direitos autorais no país. Além disso, demonstrar como as disputas pelo acesso ao conhecimento, aliadas à conjuntura internacional, retiram, momentaneamente, a lei de direitos autorais do centro da disputa, colocando em seu lugar os processos legislativos que regulamentam do uso e a gestão da rede mundial de computadores

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / This thesis is part of the Research Line Public Policy and School Practice of the Graduate Program in Education of the Federal University of Santa Maria. It is a multiple case study that aimed to identify, interpret and understand the organization of continuous training of public school teachers policies of high school in the cities of Cascavel \ PR and Santa Maria \ RS. In seeking the context of these practices analyzes focused on the Educational Development Program Paraná \ PDE / PR and continuing training policies for teachers resulting from the Proposed Restructuring Curriculum of Secondary Education of the State Secretariat of Rio Grande do Sul Education (SEDUC / RS). The analysis left of the relationship between changes in the workplace that show in different professional sectors, continuing education as a requirement to meet the needs of society governed by the intellectualization of the productive, social and cultural space. In education, continuing education of teachers has been put as essential for professional advancement, this because the decentralized education policy, due to the new organization of the now relaxed and collective work, put on the teacher responsible for the training subjects capable of understanding and enter critically in society. For the development of research used the case study (Yin, 2010; GIL, 2009) and as techniques for data collection semi-structured interviews (CHIZZOTTI, 2012; FLICK, 2009) with teachers and students of state schools as well as analysis technical and official documents of states and schools, laws, plans, courses and continuing education programs sponsored by the Secretaries of State for Education and schools. The main thesis that the proposals for continuing education practice teachers in the public education state networks are consolidating School policies in accordance with the proposals of each State has been confirmed, although in practical terms showed detachment of teachers and students of specific training in the perspective of integrated and polytechnic education. The study made it clear that the inclusion of continuing education programs for teachers contributes to the expansion of production experience and theoretical and methodological knowledge in the subjects that act; He highlighted the contradictions of scientific development that separates science from the workplace and the importance of enhancing teachers' knowledge about the principles underlying the proposed curriculum that takes work, culture, science and technology. / Esta tese integra a Linha de Pesquisa Políticas Públicas e Práticas Escolares do Programa de Pós-Graduação em Educação da Universidade Federal de Santa Maria. Trata-se de um estudo de caso múltiplo que objetivou conhecer, interpretar e compreender a organização das políticas de formação continuada de professores da escola pública estadual do Ensino Médio nos municípios de CascavelPR e Santa MariaRS. Ao buscar a contextualização destas práticas as análises incidiram sobre o Programa de Desenvolvimento Educacional do ParanáPDE/PR e as políticas de formação continuada de professores decorrentes da Proposta de Reestruturação Curricular do Ensino Médio da Secretaria Estadual de Educação do Rio Grande do Sul (SEDUC/RS). A análise partiu das relações existentes entre as mudanças no mundo do trabalho que evidenciam em diferentes setores profissionais, a formação continuada como requisito para atender aos anseios da sociedade regida pela intelectualização do espaço produtivo, social e cultural. No campo educacional, a formação continuada de professores tem sido colocada como imprescindível para o avanço profissional, isto porque as políticas educacionais descentralizadas, fruto da nova organização do trabalho agora flexibilizado e coletivo, colocam no professor a responsabilidade de formar sujeitos capazes de compreender e se inserir criticamente na sociedade. Para o desenvolvimento da pesquisa utilizou-se o estudo de caso (YIN, 2010; GIL, 2009) e como técnicas para a coleta de dados entrevistas semiestruturadas (CHIZZOTTI, 2012; FLICK, 2009) com professores e alunos de escolas estaduais além de análise de documentos técnicos e oficiais dos estados e das escolas, legislações, planos, cursos e programas de formação continuada promovida pelas Secretarias de Estado da Educação e pelas escolas. A tese principal de que as propostas de formação continuada de professores em prática nas redes estaduais públicas de ensino estão consolidando as políticas de Ensino Médio em acordo com as propostas de cada Estado foi confirmada, ainda que no plano prático demonstrasse o distanciamento de professores e de estudantes das especificidades da formação na perspectiva da educação integrada e politécnica. O estudo tornou evidente que a inserção de programas de formação continuada para os professores contribui para a ampliação de experiências e produção de conhecimentos teórico-metodológicos nas disciplinas que atuam; ressaltou as contradições do desenvolvimento científico que separa a ciência do mundo do trabalho, bem como a relevância de potencializar o conhecimento dos professores acerca dos princípios que fundamentam a proposta curricular que assume o trabalho, a cultura, a ciência e a tecnologia.

No description available.

Linear time-invariant (LTI) dynamic networks are described by their dynamical structure function, and generally, they have many possible state space realizations. This work characterizes the necessary and sufficient conditions on a state transformation that preserves the dynamical structure function, thereby generating the entire set of realizations of a given order for a specific dynamic network.