The concept of backward stochastic differential equation (BSDE) was initially brought up by Bismut when studying the stochastic optimal control problem. And it has been applied to describe various problems particularly to those in finance. After the fundamental work by Pardoux and Peng who proved the well-posedness of the nonlinear BSDE, the BSDE has been investigated intensively for both theoretical and practical purposes. In this thesis, we are concerned with a class of numerical methods for solving BSDEs, especially the one proposed by Zhao et al.. For this method, the convergence theory of the semi-discrete scheme (the scheme that discretizes the equation only in time) was already established, we shall further provide the analysis for the fully discrete scheme (the scheme that discretizes in both time and space). Moreover, using the BSDE as the adjoint equation, we shall construct the numerical method for solving the stochastic optimal control problem. We will discuss the situation when the control is deterministic as well as when the control is feedback.

This thesis concerns the construction of a detailed, compartmental population balance model for twin-screw granulation using the stochastic weighted particle method. A number of new particle mechanisms are introduced and existing mechanisms augmented including immersion nucleation, coagulation, breakage, consolidation, liquid penetration, primary particle layering and transport. The model’s predictive power is assessed over a range of liquid-solid mass feed ratios using existing experimental data and is demonstrated to qualitatively capture key experimental trends in the physical characteristic of the granular product. As part of the model development process, a number of numerical techniques for the stochastic weighed method are constructed in order to efficiently solve the population balance model. This includes a new stochastic implementation of the immersion nucleation mechanism and a variable weighted inception algorithm that dramatically reduces the number of computational particles (and hence computational power) required to solve the model. Optimum operating values for free numerical parameters and the general convergence properties of the complete simulation algorithm are investigated in depth. The model is further refined though the use of distinct primary particle and aggregate population balances, which are coupled to simulate the complete granular system. The nature of this coupling permits the inclusion of otherwise computational prohibitive mechanisms, such as primary particle layering, into the process description. A new methodology for assigning representative residence times to simulation compartments, based on screw geometry, is presented. This residence time methodology is used in conjunction with the coupled population balance framework to model twin-screw systems with a number of different screw configurations. The refined model is shown to capture key trends attributed to screw element geometry, in particular, the ability of kneading elements to distribute liquid across the granular mass.

abstract: Recently, the location of the nodes in wireless networks has been modeled as point processes. In this dissertation, various scenarios of wireless communications in large-scale networks modeled as point processes are considered. The first part of the dissertation considers signal reception and detection problems with symmetric alpha stable noise which is from an interfering network modeled as a Poisson point process. For the signal reception problem, the performance of space-time coding (STC) over fading channels with alpha stable noise is studied. We derive pairwise error probability (PEP) of orthogonal STCs. For general STCs, we propose a maximum-likelihood (ML) receiver, and its approximation. The resulting asymptotically optimal receiver (AOR) does not depend on noise parameters and is computationally simple, and close to the ML performance. Then, signal detection in coexisting wireless sensor networks (WSNs) is considered. We define a binary hypothesis testing problem for the signal detection in coexisting WSNs. For the problem, we introduce the ML detector and simpler alternatives. The proposed mixed-fractional lower order moment (FLOM) detector is computationally simple and close to the ML performance. Stochastic orders are binary relations defined on probability. The second part of the dissertation introduces stochastic ordering of interferences in large-scale networks modeled as point processes. Since closed-form results for the interference distributions for such networks are only available in limited cases, it is of interest to compare network interferences using stochastic. In this dissertation, conditions on the fading distribution and path-loss model are given to establish stochastic ordering between interferences. Moreover, Laplace functional (LF) ordering is defined between point processes and applied for comparing interference. Then, the LF orderings of general classes of point processes are introduced. It is also shown that the LF ordering is preserved when independent operations such as marking, thinning, random translation, and superposition are applied. The LF ordering of point processes is a useful tool for comparing spatial deployments of wireless networks and can be used to establish comparisons of several performance metrics such as coverage probability, achievable rate, and resource allocation even when closed form expressions for such metrics are unavailable. / Dissertation/Thesis / Ph.D. Electrical Engineering 2014

It has been shown that backward doubly stochastic differential equations (BDSDEs) provide a probabilistic representation for a certain class of nonlinear parabolic stochastic partial differential equations (SPDEs). It has also been shown that the solution of a BDSDE with Lipschitz coefficients can be approximated by first discretizing time and then calculating a sequence of conditional expectations. Given fixed points in time and space, this approximation has been shown to converge in mean square. In this thesis, we investigate the approximation of solutions of BDSDEs with coefficients that are measurable in time and space using a time discretization scheme with a view towards applications to SPDEs. To achieve this, we require the underlying forward diffusion to have smooth coefficients and we consider convergence in a norm which includes a weighted spatial integral. This combination of smoother forward coefficients and weaker norm allows the use of an equivalence of norms result which is key to our approach. We additionally take a brief look at the approximation of solutions of a class of infinite horizon BDSDEs with a view towards approximating stationary solutions of SPDEs. Whilst we remain agnostic with regards to the implementation of our discretization scheme, our scheme should be amenable to a Monte Carlo simulation based approach. If this is the case, we propose that in addition to being attractive from a performance perspective in higher dimensions, such an approach has a potential advantage when considering measurable coefficients. Specifically, since we only discretize time and effectively rely on simulations of the underlying forward diffusion to explore space, we are potentially less vulnerable to systematically overestimating or underestimating the effects of coefficients with spatial discontinuities than alternative approaches such as finite difference or finite element schemes that do discretize space. Another advantage of the BDSDE approach is that it is possible to derive an upper bound on the error of our method for a fairly broad class of conditions in a single analysis. Furthermore, our conditions seem more general in some respects than is typically considered in the SPDE literature.

Let H be a separable Hilbert space. Suppose (Ω, F, Ft, P) is a complete stochastic basis with a right continuous filtration and {Wt,t ∈ R} is an H-valued cylindrical Brownian motion with respect to {Ω, F, Ft, P). U(t, s) denotes an almost strong evolution operator generated by a family of unbounded closed linear operators on H. Consider the semilinear stochastic integral equation [formula omitted] where • f is of monotone type, i.e., ft(.) = f(t, w,.) : H → H is semimonotone, demicon-tinuous, uniformly bounded, and for each x ∈ H, ft(x) is a stochastic process which satisfies certain measurability conditions. • gs(.) is a uniformly-Lipschitz predictable functional with values in the space of Hilbert-Schmidt operators on H. • Vt is a cadlag adapted process with values in H. • X₀ is a random variable. We obtain existence, uniqueness, boundedness of the solution of this equation. We show the solution of this equation changes continuously when one or all of X₀, f, g, and V are varied. We apply this result to find stationary solutions of certain equations, and to study the associated large deviation principles. Let {Zt,t ∈ R} be an H-valued semimartingale. We prove an Ito-type inequality and a Burkholder-type inequality for stochastic convolution [formula omitted]. These are the main tools for our study of the above stochastic integral equation. / Science, Faculty of / Mathematics, Department of / Graduate

Orientador: Rogério Rosenfeld / Banca: Mario José de Oliveira / Banca: Marcos Eugênio da Silva / Resumo: Neste trabalho estudamos a calibração de opções de câmbio no mercado brasileiro utilizando o processo estocástico proposto por Heston [Heston, 1993], como uma alternativa ao modelo de apreçamento de Black e Scholes [Black e Scholes,1973], onde as volatilidades implícitas de opções para diferentes preços de exercícios e prazos são incorporadas ad hoc. Comparamos dois métodos de apreçamento: o método de Carr e Madan [Carr e Madan, 1999], que emprega transfomada rápida de Fourier e função característica, e expansão assintótica para baixos valores de volatilidade da variância. Com a nalidade de analisar o domínio de aplicabilidade deste método, selecionamos períodos de alta volatilidade no mercado, correspondente à crise subprime de 2008, e baixa volatilidade, correspondente ao período subsequente. Adicionalmente, estudamos a incorporação de swaps de variância para melhorar a calibração do modelo / Abstract: In this work we study the calibration of forex call options in the Brazilian market using the stochastic process proposed by Heston [Heston, 1993], as an alternative to the Black and Scholes [Black e Scholes,1973] pricing model, in which the implied option volatilities related to di erent strikes and maturities are incorporated in an ad hoc manner. We compare two pricing methods: one from Carr and Madan [Carr e Madan, 1999], which uses fast Fourier transform and characteristic function, and asymptotic expantion for low values of the volatility of variance. To analyze the applicability of this method, we select periods of high volatility in the market, related to the subprime crisis of 2008, and of low volatility, correspondent to the following period. In addition, we study the use of variance swaps to improve the calibration of the model / Mestre

No description available.

<p>Stochastic Gradient Descent (SGD) is a widely-used iterative algorithm for solving stochastic optimization problems for a smooth (and possibly non-convex) objective function via queries from a first-order stochastic oracle.</p> <p>In this dissertation, we critically examine SGD's choice of executing a single step as opposed to multiple steps between subsample updates. Our investigation leads naturally to generalizing SG into Retrospective Approximation (RA) where, during each iteration, a deterministic solver executes possibly multiple steps on a subsampled deterministic problem and stops when further solving is deemed unnecessary from the standpoint of statistical efficiency. RA thus leverages what is appealing for implementation -- during each iteration, a solver, e.g., L-BFGS with backtracking line search is used, as is, and the subsampled objected function is solved only to the extent necessary. We develop a complete theory using relative error of the observed gradients as the principal object, demonstrating that almost sure and L1 consistency of RA are preserved under especially weak conditions when sample sizes are increased at appropriate rates. We also characterize the iteration and oracle complexity (for linear and sub-linear solvers) of RA, and identify two practical termination criteria, one of which we show leads to optimal complexity rates. The message from extensive numerical experiments is that the ability of RA to incorporate existing second-order deterministic solvers in a strategic manner is useful both in terms of algorithmic trajectory as well as from the standpoint of  dispensing with hyper-parameter tuning.</p>

The Frank-Wolfe (FW) optimization algorithm, due to its projection free property, has gained popularity in recent years with typical application within the field of machine learning. In the stochastic setting, it is still relatively understudied in comparison to the more expensive projected method of Stochastic Gradient Descent (SGD). We propose a novel Stochastic Frank-Wolfe (SFW) algorithm, inspired by SGDo where random sampling is considered without replacement. In analogy to this abbreviation, we call the proposed algorithm SFWo. We consider a convex setting of gradient Lipschitz (smooth) functions over compact domains. Depending on experiment design (LASSO, Matrix Sensing, Matrix Completion), the SFWo algorithm, exhibits a faster or matching empirical convergence, and a tendency of bounded suboptimality by the precursor SFW. Benchmarks on both synthetic and real world data display that SFWo improves on the number of stochastic gradient evaluations needed to achieve the same guarantee as SFW. / Intresset för Frank-Wolfes (FW) optimeringsalgoritm har tack vare dess projektionsfria egenskap ökat de senaste åren med typisk tillämpning inom området maskininlärning. I sitt stokastiska uförande är den fortfarande relativt understuderad i jämförelse med den dyrare projicerande metoden Stochastic Gradient Descent (SGD). Vi föreslår en ny Stochastic Frank-Wolfe(SFW) algoritm, inspirerad av SGDo där slumpmässigt urval görs utan återläggning. I analogi med denna förkortning så kallar vi den föreslagna algoritmen SFWo. Vi betraktar en konvex miljö och gradient Lipschitz kontinuerliga (släta) funktioner över kompakta definitionsmängder. Beroende på experimentdesign (LASSO, Matrix Sensing, Matrix Completion) så visar den föreslagna algoritmen SFWo på en snabbare eller matchande empirisk konvergens och tenderar vara begränsad i suboptimalitet av föregångaren SFW. Prestandajämförelser på både syntetisk och verklig data visar att SFWo förbättrarantalet stokastiska gradientevalueringar som behövs för att uppnå samma garanti som för SFW.

When structural optimization is performed via an iterative solution technique, it is possible to integrate the analysis and design iterations, in an integrated analysis and design procedure. The present work seeks to apply an integrated analysis and design approach in reliability based optimization, when a safety index approach is used. Two variants of the new approach are presented. Both of them are based on partially converged solution of the optimization procedure. The safety index approach employed allows us to use semi-analytical formulas to calculate the sensitivity derivatives of the safety constraints. The new approach is applied to the design of a simple structure. Both methods are robust to a satisfactory degree. The results are compared to those obtained by the safety index approach without integrating the analysis and design processes. The new methods substantially reduce the computational cost of optimization, which indicates that integrated analysis and design has the potential of removing a major obstacle, which is the excessive computational cost, in applying stochastic optimization to real life structural design. / Master of Science