Optimal policies for managing cellular networks (mobile communications)

Sridhar, Suresh Vanderbilt University 1996 국내박사

BitC: A safe systems programming language

Sridhar, Swaroop The Johns Hopkins University 2009 해외박사(DDOD)

소속기관이 구독 중이 아닌 경우 오후 4시부터 익일 오전 9시까지 원문보기가 가능합니다.

Modern programming languages support the development of correct, high-performance application software by providing features such as static type safety, sound and complete type inference, higher-order functions, polymorphism, and mathematically well-founded semantics. Systems software, such as device drivers and operating system kernels, imposes the unique requirements of prescriptive data structure representation and generic mutability support on languages used for its development. Modern programming languages---particularly those that support type inference---do not support complete mutability and prescriptive data structure representation. In consequence, systems codes continue to be built in languages such as C and C++ without the benefits of modern programming language principles and tools. This dissertation presents BitC, the first language to support principal polymorphic type inference in the presence of C-style representation control and mutability. BitC's type system features a new mutability model that combines unboxed types with a consistent typing of mutability. The type system is provably sound, supports polymorphism, and eliminates the need for alias analysis to determine the immutability of a location. A sound and complete inference algorithm for this type system, which automatically infers both mutability and polymorphism, is presented. The dissertation presents an evaluation of BitC's effectiveness as a systems programming language with respect to expressiveness as well as performance. It demonstrates that key data-structures, algorithms and programming idioms used in systems programming can be safely written in BitC. Early measurements on performance critical programs (ex: SHA1 hash computation) suggest that the performance of these routines written BitC is comparable to their native C implementations.

Mechanics of Transient Networks: an Interplay of Dynamics and Microstructure

Sridhar, Shankar Lalitha University of Colorado at Boulder ProQuest Dissert 2020 해외박사(DDOD)

소속기관이 구독 중이 아닌 경우 오후 4시부터 익일 오전 9시까지 원문보기가 가능합니다.

The second law of thermodynamics states that the entropy of a system always increases with time. In this sense, all materials are transient due to the irreversibility of natural processes. Even solids like rocks are transient when talking about timescales spanning millennia. However, in terms of measuring their properties, one can assume that these materials have a "static" microstructure that is in equilibrium. This process is applicable for most hard solids where the internal arrangement of atoms and molecules do not undergo much change over short times. The same cannot be said of soft matter like polymers which are susceptible to structural change due to forces on the scale of thermal fluctuations. Biological systems rely on polymers like proteins and DNA for achieving most of the key functions of life including growth, proliferation, locomotion, self-repair etc. There is a wide variety of polymeric materials like collagen in animal tissues, the water-based hydrogels, and liquid crystals, whose underlying network structure is transient in that it can remodel and grow, degrade over time or re-organize in response to stimuli like heat or light. At the core of these materials, there is an interplay of the structural components of the polymer network and time-dependent processes that re-organize these components. Understanding this interplay through theory and modelling can deepen our understanding of biological machinery and lead the way towards smart synthetic counterparts. In this dissertation, I present a largely theoretical body of work on modelling transient networks with a bottom-up approach starting from the interactions of polymers chains to scale up to macroscopic mechanical behavior. Each chapter of this dissertation is focused on studying the characteristics that emerge for polymers networks with different properties and structures. There is a strong relevance to biological materials in most of the work presented here like cartilage growth (chapters 2 and 3), macromolecular transport into the cell nucleus (chapter 4), expansive growth of fungal cells (chapter 5) and actin networks (chapter 6); there is also significant discussions on synthetic polymers like hydrogels (chapters 2 and 3), supramolecular networks (chapter 4) and liquid crystal polymers (chapter 7).

Postsecondary enrollment effects of merit-based financial aid: Evidence from Georgia's HOPE Scholarship Program

Sridhar, Deepa Jayanth University of Georgia 2001 해외박사(DDOD)

소속기관이 구독 중이 아닌 경우 오후 4시부터 익일 오전 9시까지 원문보기가 가능합니다.

Traditionally, state and federal postsecondary student financial aid programs have been need-based. On the other hand, large-scale, government-sponsored, merit aid programs are a recent phenomenon. A prominent example is Georgia's HOPE (“Helping Outstanding Pupils Educationally”) Scholarship Program initiated in 1993, funded by a state lottery. This study analyzes the policy shift from need-based to merit based aid, as exemplified by the HOPE Scholarship, and its effects on college attendance. For meritorious Georgia students, HOPE pays the entire tuition at any in-state, 4-year or 2-year public college and to a lesser extent covers costs for students attending Georgia's private schools. Income restrictions, once present, have now been eliminated. Between 1993 and 2000 the HOPE program has subsidized over a half million students in excess of $1 billion. Its success has inspired at least a dozen other states to initiate HOPE-like programs. The natural-experiment feature of HOPE is exploited by contrasting enrollment rates in Georgia with those in a set of “control-group” states, to empirically estimate the effect of HOPE on college freshmen enrollments, by institution type and by race using state-level freshman data from the Integrated Postsecondary Education Data System (IPEDS) between 1988 and 1997 for 15 southeastern states including Georgia. The following is a summary of the main findings. (1) The overall college freshman enrollment rate in Georgia increased between 6 and 11 percent (depending on the specification chosen) due to the HOPE Scholarship between 1993 and 1997. (2) Freshman enrollment rates increased by 11 to 16 percent in all 4 year institutions; HOPE effects observed at 2-year institutions are extremely fragile. These results support the view that HOPE, unlike need-based aid, has primarily served to influence college choice, rather than increase access. (3) These effects are largest at public, 4year institutions for blacks (21 to 24 percent compared to 6 percent for whites), and at private, 4-year schools for whites (12 percent; roughly the same for blacks). (4) Finally, preliminary evidence suggest that freshman student quality proxied by SAT verbal and math percentiles, increased both at Georgia's flagship public university The University of Georgia, and at its lower-tier, state universities, post HOPE.

Effects of bibliotherapy+ on text comprehension, reading attitude, and self-concept in third and fourth grade students with attention difficulties

Sridhar, Dheepa The University of Texas at Austin 2000 해외박사(DDOD)

소속기관이 구독 중이 아닌 경우 오후 4시부터 익일 오전 9시까지 원문보기가 가능합니다.

Bibliotherapy has traditionally been defined as the use of books to help people gain insight into their behavior. Despite the use of bibliotherapy in many fields, its use in education has not been widely examined. This study investigated the effectiveness of modified bibliotherapy (bibliotherapy+) in improving text comprehension, reading attitude, and self-concept in third and fourth grade students with attention difficulties. Participants from classrooms in two comparable schools were randomly assigned to either an experimental or a control group. The intervention spanned 12 weeks with three, 45-minute sessions per week. After the conclusion of the intervention, all students in each group were evaluated on measures of reading comprehension, reading attitude, and self-concept. The effectiveness of bibliotherapy+ was examined using a repeated measure design. Results indicate the effectiveness of bibliotherapy+ in enhancing text comprehension. However, bibliotherapy+ had no effect on reading attitude and resulted in a decrease in self-concept scores.

Mixed-integer programming approaches for some non-convex and combinatorial optimization problems

Sridhar, Srikrishna The University of Wisconsin - Madison 2014 해외박사(DDOD)

소속기관이 구독 중이 아닌 경우 오후 4시부터 익일 오전 9시까지 원문보기가 가능합니다.

In this dissertation we study several nonconvex and combinatorial optimization problems with applications in production planning, machine learning, advertising, statistics, and computer vision. The common theme is the use of algorithmic and modelling techniques from mixed-integer programming (MIP) which include formulation strengthening, decomposition, and linear programming (LP) rounding. We first consider MIP formulations for piecewise linear functions (PLFs) that are evaluated when an indicator variable is turned on. We describe modifications to standard MIP formulations for PLFs with desirable theoretical properties and superior computational performance in this context. Next, we consider a production planning problem where the production process creates a mixture of desirable products and undesirable byproducts. In this production process, at any point in time the fraction of the mixture that is an undesirable byproduct increases monotonically as a function of the cumulative mixture production up to that time. The mathematical formulation of this continuous-time problem is nonconvex. We present a discrete time mixed-integer nonlinear programming (MINLP) formulation that exploits the increasing nature of the byproduct ratio function. We demonstrate that this new formulation is more accurate than a previously proposed MINLP formulation. We describe three different mixedinteger linear programming (MIP) approximation and relaxation models of this nonconvex MINLP, and derive modifications that strengthen the LP-relaxations of these models. We provide computational experiments that demonstrate that the proposed formulation is more accurate than the previous formulation, and that the strengthened MIP approximation and relaxation models can be used to obtain near-optimal solutions for large instances of this nonconvex MINLP. We then study production planning problems in the presence of realistic business rules like taxes, tariffs, and royalties. We propose two different solution techniques. The first is a MIP formulation while the second is a search algorithm based on a novel continuous domain formulation. We then discuss decomposition methods to compute bounds on the optimal solution. Our computational experiments demonstrate the impact of our formulations, solution techniques, and algorithms on a sample application problem. Finally, we study three classes of combinatorial optimization problems: set packing, set covering, and multiway-cut. Near-optimal solutions of these combinatorial problems can be computed by rounding the solution of an LP. We show that one can recover solutions of comparable quality by rounding an approximate LP solution instead of an exact one. These approximate LP solutions can be computed efficiently by solving a quadratic-penalty formulation of the LP using a parallel stochastic coordinate descent method. We derive worst-case runtime and solution quality guarantees of this scheme using novel perturbation and convergence analyses. Our experiments demonstrate that on these combinatorial problems our rounding scheme is up to an order of magnitude faster than Cplex (a commercial LP solver) while producing solutions of similar quality.

Inference of Cascades and Correlated Networks

Sridhar, Anirudh Princeton University ProQuest Dissertations & Thes 2023 해외박사(DDOD)

소속기관이 구독 중이 아닌 경우 오후 4시부터 익일 오전 9시까지 원문보기가 가능합니다.

This thesis makes fundamental contributions to a few statistical inference tasks on networks, with a focus on information-theoretic characterizations. In the first part of this thesis, we study the problem of localizing a network cascade from noisy, real-time measurements of its spread (i.e., through error-prone diagnostic tests). Our objective is to design algorithms that can estimate the cascade source as fast as possible, so that the impact of the cascade on the network can be mitigated. We design estimation procedures from Bayesian and minimax perspectives. In the Bayesian setting, we propose an estimator which observes samples until the error of the Bayes-optimal estimator falls below a threshold. In the minimax setting, we devise a novel multihypothesis sequential probability ratio test (MSPRT) for source estimation. When estimating simple cascades in trees and lattices, we show that both methods are optimal, in the sense that no other algorithm can accurately estimate the source with a substantially smaller number of samples. Finally, we discuss how our methods may be extended to estimate realistic cascades in generic networks.In the second part of this thesis, we study graph matching and community recovery in networks with correlated structure. First, we derive the precise information-theoretic threshold for fully recovering the latent vertex correspondence between two edge-correlated stochastic block models - a task known as exact graph matching. We then characterize the information-theoretic landscape of community recovery in correlated stochastic block models, which requires a delicate interplay between graph matching and community recovery algorithms. In particular, we uncover and characterize a region of the parameter space where exact community recovery is possible using multiple correlated graphs, even though (1) this is information-theoretically impossible using a single graph and (2) exact graph matching is also information-theoretically impossible. In this regime, we develop a novel algorithm that carefully synthesizes community recovery and graph matching algorithms.

Limit Shapes in the Six Vertex Model

Sridhar, Ananth University of California, Berkeley ProQuest Disser 2016 해외박사(DDOD)

소속기관이 구독 중이 아닌 경우 오후 4시부터 익일 오전 9시까지 원문보기가 가능합니다.

The purpose of this thesis is to present some new results about the six-vertex model and dimer model, andi in particular some aspects of the limit shapes that form in the thermodynamic limit and the partial differential equations that arise in their study. Chapter 1 provides a detailed introduction to limit shape phenomena and its mathematics. Chapter 2 reviews basic facts about the six-vertex model and dimer model and their properties in the thermodynamic limit. Chapter 3 investigates the role of integrability of the six vertex model in the formation of limit shapes. By integrability we mean, on one hand the discrete integrability of the six-vertex model in the sense of commutative families of transfer matrices, and on the other hand the integrability in the continuum limit in the Liouville sense of Poisson commutative families of functions (integrals of motion) on the phase space. The main result is to show that the partial differential equations describing the limit shape have an infinite number of conserved quantities. Chapter 4 explains the results of the previous chapter in a more general setting. The main result is a straightforward theorem in Hamiltonian mechanics, which gives conditions for the Poisson commutativity of two Hamiltonian functions in terms of their principal action functions. It suggests a generalization of our previous results to other integrable lattice models, for example, the generalizations of the six-vertex model related to other Lie algebras. Chapter 5 studies the six-vertex model for a special class of weights called the stochastic six-vertex model. for which the six vertex model is closely related to interacting particle models. The main result of this chapter is the derivation for the partial differential equations for the height function of the stochastic 6-vertex model on the cylinder. Chapter 6 studies models at the intersection of discrete geometry and statistical mechanics. The main result is the asymptotic expansion of the discrete Laplacian that approximate Laplace operators induced by a Riemannian metric. They arise as a discretization of the of the Gaussian free field and are closely related to Dimer models that can be regarded as a discretization of Dirac operators in a background metric.

Essays on Innovation

Srinivasan, Sridhar Northwestern University ProQuest Dissertations & T 2019 해외박사(DDOD)

소속기관이 구독 중이 아닌 경우 오후 4시부터 익일 오전 9시까지 원문보기가 가능합니다.

This dissertation explores two factors that affect innovation. The first factor studied is extrinsic, namely, the legal environment. The second factor is a factor of input into innovation - culture of the inventors.The first chapter studies the impact of a weakened patent environment on research investments. Departing from the bulk of the literature, this chapter focuses on frivolous patenting among productive firms as opposed to frivolous patenting by patent trolls. A simple theory is proposed to model strategic behavior among productive firms in an environment that allows for frivolous patenting.When the law is changed to disallow this arms race, the model makes predictions on how the firms respond as a function of their exposure to the law. Specifically and somewhat surprisingly, the model predicts that the most exposed firms continue patenting similar to patent trolls since these firms extract rents due to the liberal patenting environment. Conditional on being exposed, the least exposed firms behave as one would expect non-patent trolls to behave and respond by dramatically reducing their patenting activity. Further, the model predicts scaled R&D investments respond in the shape of an inverted U-curve with respect to level of exposure to the law change.Both these predictions are verified in the data. These results lend credence to the hypothesis that productive firms file frivolous patents similar to patent trolls and that patent law can, on occasion, be too liberal in terms of allowing patentable subject matter.The second chapter focuses on one of the inputs into the innovation process - the inventors themselves. Using recently available machine learning techniques, ethnicities are inferred from the names of inventors on patents. Using the ethnicity of the inventors as proxy for culture, this work examines whether culture affects innovation.The first finding is that culture does play a role in influencing quality of innovation measures through citations, stock market value and a text based metric of impact and novelty. This finding shows that ethnicity is significant when predicting quality outcomes for patents.Secondly, when considering the list of inventors on a patent as a team, this chapter shows that ethnically diverse teams have higher quality patents. Furthermore, the gain in quality is not just from an increase from a homogeneous team to an non-homogeneous team but also from moving to a less diverse team to a more diverse team.

Gene Correction of Limb Girdle Muscular Dystrophy Type 2A Patient-Specific Induced Pluripotent Stem Cells

Selvaraj, Sridhar ProQuest Dissertations & Theses University of Minn 2019 해외박사(DDOD)

소속기관이 구독 중이 아닌 경우 오후 4시부터 익일 오전 9시까지 원문보기가 가능합니다.

Targeted differentiation of pluripotent stem (PS) cells into myotubes enables in vitro disease modeling of skeletal muscle diseases. Although various protocols achieve myogenic differentiation in vitro, resulting myotubes invariably display an embryonic identity. This is a major hurdle for accurately recapitulating disease phenotypes in vitro, as disease typically does not manifest in the embryonic muscle, but at more mature stages. To address this problem, we identified four factors from a small molecule screen whose combinatorial treatment resulted in myotubes with enhanced maturation, as shown by increased expression of fetal, neonatal and adult myosin heavy-chain isoforms. These molecular changes were confirmed by global chromatin accessibility and transcriptome studies. Importantly, we also observed this maturation in three-dimensional muscle bundles, which displayed improved in vitro contractile force generation in response to electrical stimulus. Thus, we established a model for in vitro muscle maturation from PS cells.We applied this maturation model for in vitro validation of Calpain 3 (CAPN3) protein expression. CAPN3 mutations are associated with Limb Girdle Muscular Dystrophy type 2A (LGMD2A), which is an incurable autosomal recessive disorder that results in muscle wasting and loss of ambulation. Using a gene knock-in approach, here we applied CRISPR-Cas9 mediated genome editing to induced pluripotent stem (iPS) cells from three LGMD2A patients carrying three different CAPN3 mutations, to enable correction of mutations in the CAPN3 gene. CAPN3 protein rescue upon gene correction was validated in myotube-derivatives in vitro following the small molecule treatment. Transplantation of gene corrected LGMD2A myogenic progenitors in a novel mouse model combining immunodeficiency and lack of CAPN3 resulted in muscle engraftment and rescue of the CAPN3 mRNA. Thus, we provide here proof concept for the integration of genome editing and iPS cell technologies to develop a novel autologous cell therapy for LGMD2A.