No description available.

No description available.

Jump-diffusion models can under certain assumptions be expressed as partial integro-differential equations (PIDE). Such a PIDE typically involves a convection term and a nonlocal integral like for the here considered models of Merton and Kou. We transform the PIDE to eliminate the convection term, discretize it implicitly using finite differences and the second order backward difference formula (BDF2) on a uniform grid. The arising dense linear system is solved by an iterative method, either a splitting technique or a circulant preconditioned conjugate gradient method. Exploiting the Fast Fourier Transform (FFT) yields the solution in only $O(n\log n)$ operations and just some vectors need to be stored. Second order accuracy is obtained on the whole computational domain for Merton's model whereas for Kou's model first order is obtained on the whole computational domain and second order locally around the strike price. The solution for the PIDE with convection term can oscillate in a neighborhood of the strike price depending on the choice of parameters, whereas the solution obtained from the transformed problem is stabilized. / Master of Science

Hyperthermia is a cancer treatment modality that raises cancerous tissue to cytotoxic temperature levels for roughly 30 to 45 minutes. Hyperthermia treatment planning refers to the use of computational models to optimize the heating protocol to be used in a hyperthermia treatment. This thesis presents a method to optimize a hyperthermia treatment heating protocol. An algorithm is developed which recovers a heating protocol that will cause a desired amount of thermal damage within a region of tissue. The optimization algorithm is validated experimentally on an albumen tissue phantom. The transient temperature distribution within the region is simulated using a two-dimensional, finite-difference model of the Pennes bioheat equation. The relationship between temperature and time is integrated to produce a damage field according to two different models; Henriques'' model and the thermal dose model (Moritz and Henriques (1947)), (Sapareto and Dewey (1984)). A minimization algorithm is developed which re duces the value of an objective function based on the squared difference between an optimal and calculated damage field. Either damage model can be used in the minimization algorithm. The adjoint problem in conjunction with the conjugate gradient method is used to minimize the objective function of the control problem. The flexibility of the minimization algorithm is proven experimentally and through a variety of simulations. With regards to the validation experiment, the optimal and recovered regions of permanent thermal damage are in good agreement for each test performed. A sensitivity analysis of the finite difference and damage models shows that the experimentally-obtained extent of damage is consistently within a tolerable error range. Excellent agreement between the optimal and recovered damage fields is also found in simulations of hyperthermia treatments on perfused tissue. A simplified and complex model of the human skin were created for use within the algorithm. Minimizations using both the Henriques'' model and the thermal dose model in the objective function are performed. The Henriques'' damage model was found to be more desirable for use in the minimization algorithm than the thermal dose model because it is less computationally intensive and includes a mechanism to predict the threshold of permanent thermal damage. The performance of the minimization algorithm was not hindered by adding complexity to the skin model. The method presented here for optimizing hyperthermia treatments is shown to be robust and merits further investigation using more complicated patient models. / Master of Science

In this thesis, we develop block-decomposition (BD) methods and variants of accelerated *9gradient methods for large-scale conic programming and convex optimization, respectively. The BD methods, discussed in the first two parts of this thesis, are inexact versions of proximal-point methods applied to two-block-structured inclusion problems. The adaptive accelerated methods, presented in the last part of this thesis, can be viewed as new variants of Nesterov's optimal method. In an effort to improve their practical performance, these methods incorporate important speed-up refinements motivated by theoretical iteration-complexity bounds and our observations from extensive numerical experiments. We provide several benchmarks on various important problem classes to demonstrate the efficiency of the proposed methods compared to the most competitive ones proposed earlier in the literature. In the first part of this thesis, we consider exact BD first-order methods for solving conic semidefinite programming (SDP) problems and the more general problem that minimizes the sum of a convex differentiable function with Lipschitz continuous gradient, and two other proper closed convex (possibly, nonsmooth) functions. More specifically, these problems are reformulated as two-block monotone inclusion problems and exact BD methods, namely the ones that solve both proximal subproblems exactly, are used to solve them. In addition to being able to solve standard form conic SDP problems, the latter approach is also able to directly solve specially structured non-standard form conic programming problems without the need to add additional variables and/or constraints to bring them into standard form. Several ingredients are introduced to speed-up the BD methods in their pure form such as: adaptive (aggressive) choices of stepsizes for performing the extragradient step; and dynamic updates of scaled inner products to balance the blocks. Finally, computational results on several classes of SDPs are presented showing that the exact BD methods outperform the three most competitive codes for solving large-scale conic semidefinite programming. In the second part of this thesis, we present an inexact BD first-order method for solving standard form conic SDP problems which avoids computations of exact projections onto the manifold defined by the affine constraints and, as a result, is able to handle extra large-scale SDP instances. In this BD method, while the proximal subproblem corresponding to the first block is solved exactly, the one corresponding to the second block is solved inexactly in order to avoid finding the exact solution of a linear system corresponding to the manifolds consisting of both the primal and dual affine feasibility constraints. Our implementation uses the conjugate gradient method applied to a reduced positive definite dual linear system to obtain inexact solutions of the latter augmented primal-dual linear system. In addition, the inexact BD method incorporates a new dynamic scaling scheme that uses two scaling factors to balance three inclusions comprising the optimality conditions of the conic SDP. Finally, we present computational results showing the efficiency of our method for solving various extra large SDP instances, several of which cannot be solved by other existing methods, including some with at least two million constraints and/or fifty million non-zero coefficients in the affine constraints. In the last part of this thesis, we consider an adaptive accelerated gradient method for a general class of convex optimization problems. More specifically, we present a new accelerated variant of Nesterov's optimal method in which certain acceleration parameters are adaptively (and aggressively) chosen so as to: preserve the theoretical iteration-complexity of the original method; and substantially improve its practical performance in comparison to the other existing variants. Computational results are presented to demonstrate that the proposed adaptive accelerated method performs quite well compared to other variants proposed earlier in the literature.

<div>The objective of this doctoral research was to develop a versatile inverse heat transfer approach, that would enable the solution of small scale problems present in microelectronics, as well as the analysis of the complex heat flux in turbines. An inverse method is a mathematical approach which allows the resolution of problems starting from the solution. In a direct problem, the boundary conditions are given, and using the governing physics principles and equations you can calculate the solution or physical effect. In an inverse method, the solution is provided and through the physical equations, the boundary conditions can be determined. Therefore, the inverse method applied to heat transfer means that we know the variation of temperature (effect) over time and space. With the temperature input, the geometry, thermal properties of the test article and the heat diffusion equation, we can compute the spatially- and temporally-varying heat flux that generated the temperature map.</div><div><br></div><div>This doctoral dissertation develops two inverse methodologies: (1) an optimization methodology based on the conjugate gradient method and (2) a function specification method combined with a regularization technique, which is less robust but much faster. We implement these methodologies with commercial codes for solving conductive heat transfer with COMSOL and for conjugate heat transfer with ANSYS Fluent.</div><div><br></div><div>The goal is not only the development of the methods but also the validation of the techniques in two different fields with a common purpose: quantifying heat dissipation. The inverse methods were applied in the micro-scale to the dissipation of heat in microelectronics and in the macro-scale to the gas turbine engines.<br></div><div><br></div><div>In microelectronics, we performed numerical and experimental studies of the two developed inverse methodologies. The intent was to predict where heat is being dissipated and localized hot spots inside of the chip from limited measurements of the temperature outside of the chip. Here, infrared thermography of the chip surface is the input to the inverse methods leveraging thermal model of the chip. Furthermore, we combined the inverse methodology with a Kriging interpolation technique with genetic algorithm optimization to optimize the location and number of the temperature sensors inside of the chip required to accurately predict the thermal behavior of the microchip at each moment of time and everywhere.<br></div><div><br></div><div>In the application for gas turbine engines, the inverse method can be useful to detect or predict the conditions inside of the turbine by taking measurements in the outer casing. Therefore, the objective is the experimental validation of the technique in a wind tunnel especially designed with optical access for non-contact measurement techniques. We measured the temperature of the outer casing of the turbine rotor with an infrared camera and surface temperature sensors and this information is the input of the two methodologies developed in order to predict which the heat flux through the turbine casing. A new facility, specifically, an annular turbine cascade, was designed to be able to measure the relative frame of the rotor from the absolute frame. In order to get valuable data of the heat flux in a real engine, we need to replicate the Mach, Reynolds, and temperature ratios between fluid and solid. Therefore, the facility can reproduce a large range of pressures and flow temperatures. Because some regions of interest are not accessible, this researchprovides a significant benefit for understanding the system performance from limited data. With inverse methods, we can measure the outside of objects and provide an accurate prediction of the behavior of the complete system. This information is relevant not only for new designs of gas turbines or microchips, but also for old designs where due to lack of prevision there are not enough sensors to monitor the thermal behavior of the studied system.<br></div><div><br></div>

Thesis (Ph. D.)--York University, 2001. Graduate Programme in Mathematics. / Typescript. Includes bibliographical references (leaves 114-119). Also available on the Internet. MODE OF ACCESS via web browser by entering the following URL: http://wwwlib.umi.com/cr/yorku/fullcit?pNQ66358.

In a series of work initiated by Nemirovsky and Yudin, and later extended by Nesterov, first-order algorithms for unconstrained minimization with optimal theoretical complexity bound have been proposed. On the other hand, conjugate gradient algorithms as one of the widely used first-order techniques suffer from the lack of a finite complexity bound. In fact their performance can possibly be quite poor. This dissertation is partially on tightening the gap between these two classes of algorithms, namely the traditional conjugate gradient methods and optimal first-order techniques. We derive conditions under which conjugate gradient methods attain the same complexity bound as in Nemirovsky-Yudin's and Nesterov's methods. Moreover, we propose a conjugate gradient-type algorithm named CGSO, for Conjugate Gradient with Subspace Optimization, achieving the optimal complexity bound with the payoff of a little extra computational cost. We extend the theory of CGSO to convex problems with linear constraints. In particular we focus on solving $l_1$-regularized least square problem, often referred to as Basis Pursuit Denoising (BPDN) problem in the optimization community. BPDN arises in many practical fields including sparse signal recovery, machine learning, and statistics. Solving BPDN is fairly challenging because the size of the involved signals can be quite large; therefore first order methods are of particular interest for these problems. We propose a quasi-Newton proximal method for solving BPDN. Our numerical results suggest that our technique is computationally effective, and can compete favourably with the other state-of-the-art solvers.

The Method of Moments (MoM) has been used extensively to solve electromagnetic problems. Its popularity is largely attributed to its adaptability to structures with various shapes and success in predicting the equivalent induced currents accurately. However, due to its dense matrix, especially for large structures, the MoM suffers from long matrix solution time and large storage requirement. In this thesis it is shown that use of wavelet basis functions result in a MoM matrix which is sparser than the one obtained by using traditional basis functions. A new wavelet system, different from the ones found in literature, is proposed. Stabilized Bi-Conjugate Gradient Method which is an iterative matrix solution method is utilized to solve the resulting sparse matrix equation. Both a one-dimensional problem with a microstrip line example and a two-dimensional problem with a rectangular patch antenna example are studied and the results are compared.

Η συνεισφορά της παρούσας διατριβής επικεντρώνεται στην ανάπτυξη και στη Μαθηματική θεμελίωση νέων μεθόδων συζυγών κλίσεων για βελτιστοποίηση χωρίς περιορισμούς και στη μελέτη νέων μεθόδων εκπαίδευσης νευρωνικών δικτύων και εφαρμογών τους. Αναπτύσσουμε δύο νέες μεθόδους βελτιστοποίησης, οι οποίες ανήκουν στην κλάση των μεθόδων συζυγών κλίσεων. Οι νέες μέθοδοι βασίζονται σε νέες εξισώσεις της τέμνουσας με ισχυρά θεωρητικά πλεονεκτήματα, όπως η προσέγγιση με μεγαλύτερη ακρίβεια της επιφάνεια της αντικειμενικής συνάρτησης. Επιπλέον, μία σημαντική ιδιότητα και των δύο προτεινόμενων μεθόδων είναι ότι εγγυώνται επαρκή μείωση ανεξάρτητα από την ακρίβεια της γραμμικής αναζήτησης, αποφεύγοντας τις συχνά αναποτελεσματικές επανεκκινήσεις. Επίσης, αποδείξαμε την ολική σύγκλιση των προτεινόμενων μεθόδων για μη κυρτές συναρτήσεις. Με βάση τα αριθμητικά μας αποτελέσματα καταλήγουμε στο συμπέρασμα ότι οι νέες μέθοδοι έχουν πολύ καλή υπολογιστική αποτελεσματικότητα, όπως και καλή ταχύτητα επίλυσης των προβλημάτων, υπερτερώντας σημαντικά των κλασικών μεθόδων συζυγών κλίσεων. Το δεύτερο μέρος της διατριβής είναι αφιερωμένο στην ανάπτυξη και στη μελέτη νέων μεθόδων εκπαίδευσης νευρωνικών δικτύων. Προτείνουμε νέες μεθόδους, οι οποίες διατηρούν τα πλεονεκτήματα των κλασικών μεθόδων συζυγών κλίσεων και εξασφαλίζουν τη δημιουργία κατευθύνσεων μείωσης αποφεύγοντας τις συχνά αναποτελεσματικές επανεκκινήσεις. Επιπλέον, αποδείξαμε ότι οι προτεινόμενες μέθοδοι συγκλίνουν ολικά για μη κυρτές συναρτήσεις. Τα αριθμητικά αποτελέσματα επαληθεύουν ότι οι προτεινόμενες μέθοδοι παρέχουν γρήγορη, σταθερότερη και πιο αξιόπιστη σύγκλιση, υπερτερώντας των κλασικών μεθόδων εκπαίδευσης. Η παρουσίαση του ερευνητικού μέρους της διατριβής ολοκληρώνεται με μία νέα μέθοδο εκπαίδευσης νευρωνικών δικτύων, η οποία βασίζεται σε μία καμπυλόγραμμη αναζήτηση. Η μέθοδος χρησιμοποιεί τη BFGS ενημέρωση ελάχιστης μνήμης για τον υπολογισμό των κατευθύνσεων μείωσης, η οποία αντλεί πληροφορία από την ιδιοσύνθεση του προσεγγιστικού Eσσιανού πίνακα, αποφεύγοντας οποιαδήποτε αποθήκευση ή παραγοντοποίηση πίνακα, έτσι ώστε η μέθοδος να μπορεί να εφαρμοστεί για την εκπαίδευση νευρωνικών δικτύων μεγάλης κλίμακας. Ο αλγόριθμος εφαρμόζεται σε προβλήματα από το πεδίο της τεχνητής νοημοσύνης και της βιοπληροφορικής καταγράφοντας πολύ καλά αποτελέσματα. Επίσης, με σκοπό την αύξηση της ικανότητας γενίκευσης των εκπαιδευόμενων δικτύων διερευνήσαμε πειραματικά και αξιολογήσαμε την εφαρμογή τεχνικών μείωσης της διάστασης δεδομένων στην απόδοση της γενίκευσης των τεχνητών νευρωνικών δικτύων σε μεγάλης κλίμακας δεδομένα βιοϊατρικής. / The contribution of this thesis focuses on the development and the Mathematical foundation of new conjugate gradient methods for unconstrained optimization and on the study of new neural network training methods and their applications. We propose two new conjugate gradient methods for unconstrained optimization. The proposed methods are based on new secant equations with strong theoretical advantages i.e. they approximate the surface of the objective function with higher accuracy. Moreover, they have the attractive property of ensuring sufficient descent independent of the accuracy of the line search, avoiding thereby the usual inefficient restarts. Further, we have established the global convergence of the proposed methods for general functions under mild conditions. Based on our numerical results we conclude that our proposed methods outperform classical conjugate gradient methods in both efficiency and robustness. The second part of the thesis is devoted on the study and development of new neural network training algorithms. More specifically, we propose some new training methods which preserve the advantages of classical conjugate gradient methods while simultaneously ensure sufficient descent using any line search, avoiding thereby the usual inefficient restarts. Moreover, we have established the global convergence of our proposed methods for general functions. Encouraging numerical experiments on famous benchmarks verify that the presented methods provide fast, stable and reliable convergence, outperforming classical training methods. Finally, the presentation of the research work of this dissertation is fulfilled with the presentation of a new curvilinear algorithm for training large neural networks which is based on the analysis of the eigenstructure of the memoryless BFGS matrices. The proposed method preserves the strong convergence properties provided by the quasi-Newton direction while simultaneously it exploits the nonconvexity of the error surface through the computation of the negative curvature direction without using any storage and matrix factorization. Our numerical experiments have shown that the proposed method outperforms other popular training methods on famous benchmarks. Furthermore, for improving the generalization capability of trained ANNs, we explore the incorporation of several dimensionality reduction techniques as a pre-processing step. To this end, we have experimentally evaluated the application of dimensional reduction techniques for increasing the generalization capability of neural network in large biomedical datasets.