Search

Near-infrared spectroscopy (NIRS) is a technique that applies light sources of multiple wavelengths within the object for non-invasive property detection. The current NIRS techniques include Continuous-Wave NIRS (CW-NIRS), Frequency-Domain NIRS (FD-NIRS), and Time-Domain NIRS (TD-NIRS). Both CW-NIRS and FD-NIRS apply constant or intensitymodulated photon-emitting light sources and photodiodes as sensors to detect the power of the light from the target in a certain time. Even though FD-NIRS uses a frequency sinusoidal function (MHz) for the modulation of the light source, which brings in more information than CW-NIRS, such as average intensity (DC), the amplitude of intensity (AC), and phase shift between the detected and emitted light, the sensitive nature of the frequency modulation and detection techniques employed makes the measurements susceptible to motion artefacts and environmental noise. These external factors can disrupt the phase and amplitude of the modulated signal, resulting in distortions that impact the accuracy of the reconstruction process and require careful consideration and mitigation strategies in the measurements. TD-NIRS applies lasers as light sources and single-photon sensors for light detection, which not only inherits the merits of CW-NIRS on the intensity measurement, but also provides up to picosecond time resolution for the average photon path length calculation and the chance to extract the absorption and scattering coefficient of the target for absolute haemoglobin concentration qualification based on the nature of the histograms. The research on TD-NIRS has been over 20 years, however, there are only few instruments in the market due to the complexity of the system and the high price, which hinders its application for clinical practices and daily care, and blocks the feedback from the actual usage. In recent years, high-resolution and high-frequency laser drivers and sources, and single-photon detectors have become more accessible and cheaper, which contributes to further research and development of TD-NIRS systems. This thesis targets to reduce the complexity and size of the TD-NIRS system while keeping the quality of spatial resolution, time resolution, and frame acquisition rate, utilising laser diodes, a Single Photon Avalanche Diode (SPAD) sensor, and a Field Programmable Gate Array (FPGA) based processing circuits. Different from previous publications, the Time-to-Digital Converter (TDC) is moved out of the SPAD and implemented in an FPGA for cost-saving and system flexibility. A novel method is applied to the TDC by encoding the states of the delay lines instead of the thermometer code used in the conventional TDCs to improve the linearity. The achieved raw Differential Non-Linearity (DNL) which is the DNL without compensation and calibration, together with the zero empty bins, to our knowledge, exceeds previously reported of the FPGA-based TDCs. The TDC is applied to the TD-NIRS system, with the achieved system Impulse Response Function (IRF) of 165 ps at the maximum repetition rate of 20 Mhz. The concept of different lasers with a sync phase shift of 2 ns to generate multi-IRFs in one histogram frame was successfully proved, providing a high-speed and real-time solution for achieving high spatial resolution.

In this thesis, we study the fi ltering problem for a partially observed jump diffusion (Zₜ)ₜɛ[ₒ,T] = (Xₜ, Yₜ)tɛ[ₒ,T] driven by Wiener processes and Poisson martingale measures, such that the signal and observation noises are correlated. We derive the fi ltering equations, describing the time evolution of the normalised conditional distribution (Pₜ(dx))tɛ[ₒ,T] and the unnormalised conditional distribution of the unobservable signal Xₜ given the observations (Yₛ)ₛɛ[ₒ,T]. We prove that if the coefficients satisfy linear growth and Lipschitz conditions in space, as well as some additional assumptions on the jump coefficients, then, if E|πₒ|ᵖLρ < ∞ for some p ≥ 2, the conditional density π = (πₜ)tɛ[ₒ,T], where πₜ = dPₜ/dx, exists and is a weakly cadlag Lp-valued process. Moreover, for an integer m ≥ 0 and p ≥ 2, we show that if we additionally impose m + 1 continuous and bounded spatial derivatives on the coefficients and if the initial conditional density E|πₒ|ᵖWρᵐ < ∞, then π is weakly cadlag as a Wρᵐ-valued process and strongly cadlag as a Wρˢ - valued process for s ɛ [0;m).

In this thesis, we investigate the Lp-solvability of a class of (possibly) degenerate stochastic integro-differential equations (SIDEs) of parabolic type, which includes the Zakai equation in nonlinear filtering for jump diffusions and the Kolmogorov equations for jump diffusions. We first study the solvability of integro-differential equations in the same type but without randomness. Then we present an Itˆo formula for the Lp-norm of jump processes having stochastic differentials in Lp-spaces, which can be used to study the solvability of SIDEs. In the last chapter, existence and uniqueness of the solutions to SIDEs are established in Bessel potential spaces.

This thesis addresses the absence of explicit pricing formulae and the complexity of proposed models (incomplete markets framework) in the area of mortality risk management requiring the application of advanced techniques from the realm of Financial Mathematics and Actuarial Science. In fact, this is a multi-essay dissertation contributing in the direction of designing and pricing mortality-linked derivatives and offering the state of art solutions to manage longevity risk. The first essay investigates the valuation of Catastrophic Mortality Bonds and, in particular, the case of the Swiss Re Mortality Bond 2003 as a primary example of this class of assets. This bond was the first Catastrophic Mortality Bond to be launched in the market and encapsulates the behaviour of a well-defined mortality index to generate payoffs for bondholders. Pricing this type of bond is a challenging task and no closed form solution exists in the literature. In my approach, we adapt the payoff of such a bond in terms of the payoff of an Asian put option and present a new methodology to derive model-independent bounds for catastrophic mortality bonds by exploiting the theory of comonotonicity. While managing catastrophic mortality risk is an upheaval task for insurers and re-insurers, the insurance industry is facing an even bigger challenge - the challenge of coping up with increased life expectancy. The recent years have witnessed unprecedented changes in mortality rate. As a result academicians and practitioners have started treating mortality in a stochastic manner. Moreover, the assumption of independence between mortality and interest rate has now been replaced by the observation that there is indeed a correlation between the two rates. Therefore, my second essay studies valuation of Guaranteed Annuity Options (GAOs) under the most generalized modeling framework where both interest rate and mortality risk are stochastic and correlated. Pricing these types of options in the correlated environment is an arduous task and a closed form solution is non-existent. In my approach, I employ the use of doubly stochastic stopping times to incorporate the randomness about the time of death and employ a suitable change of measure to facilitate the valuation of survival benefit, there by adapting the payoff of the GAO in terms of the payoff of a basket call option. I then derive general price bounds for GAOs by employing the theory of comonotonicity and the Rogers-Shi (Rogers and Shi, 1995) approach. Moreover, I suggest some `model-robust' tight bounds based on the moment generating function (m.g.f.) and characteristic function (c.f.) under the affine set up. The strength of these bounds is their computational speed which makes them indispensable for annuity providers who rely heavily on Monte Carlo simulations to calculate the fair market value of Guaranteed Annuity Options. In fact, sans Monte Carlo, the academic literature does not offer any solution for the pricing of the GAOs. I illustrate the performance of the bounds for a variety of affine processes governing the evolution of mortality and the interest rate by comparing them with the benchmark Monte Carlo estimates. Through my work, I have been able to express the payoffs of two well known modern mortality products in terms of payoffs of financial derivatives, there by filling the gaps in the literature and offering state of art techniques for pricing of these sophisticated instruments.

This thesis consists of several problems concerning numerical approximations for stochastic differential equations, and is divided into three parts. The first one is on the integrability and asymptotic stability with respect to a certain class of Lyapunov functions, and the preservation of the comparison theorem for the explicit numerical schemes. In general, those properties of the original equation can be lost after discretisation, but it will be shown that by some suitable modification of the Euler scheme they can be preserved to some extent while keeping the strong convergence rate maintained. The second part focuses on the approximation of iterated stochastic integrals, which is the essential ingredient for the construction of higher-order approximations. The coupling method is adopted for that purpose, which aims at finding a random variable whose law is easy to generate and is close to the target distribution. The last topic is motivated by the simulation of equations driven by Lévy processes, for which the main difficulty is to generalise some coupling results for the one-dimensional central limit theorem to the multi-dimensional case.

We present a new pathwise approximation method for stochastic differential equations driven by Brownian motion which does not require simulation of the stochastic integrals. The method is developed to give Wasserstein bounds O(h3/2) and O(h2) which are better than the Euler and Milstein strong error rates O(√h) and O(h) respectively, where h is the step-size. It assumes nondegeneracy of the diffusion matrix. We have used the Taylor expansion but generate an approximation to the expansion as a whole rather than generating individual terms. We replace the iterated stochastic integrals in the method by random variables with the same moments conditional on the linear term. We use a version of perturbation method and a technique from optimal transport theory to find a coupling which gives a good approximation in Lp sense. This new method is a Runge-Kutta method or so-called derivative-free method. We have implemented this new method in MATLAB. The performance of the method has been studied for degenerate matrices. We have given the details of proof for order h3/2 and the outline of the proof for order h2.

We study a new method for the strong approximate solution of stochastic differential equations using coupling and we prove order one error bounds for the new scheme in Lp space assuming the invertibility of the diffusion matrix. We introduce and implement two couplings called the exact and approximate coupling for this scheme obtaining good agreement with the theoretical bound. Also we describe a method for non-invertibility case (Combined method) and we investigate its convergence order which will give O(h3/4 √log(h)j) under some conditions. Moreover we compare the computational results for the combined method with its theoretical error bound and we have obtained a good agreement between them. In the last part of this thesis we work out the performance of the multilevel Monte Carlo method using the new scheme with the exact coupling and we compare the results with the trivial coupling for the same scheme.

We consider linear and semilinear stochastic partial differential equations that in some sense can be viewed as being at the "endpoints" of the classical variational theory by Krylov and Rozovskii [25]. In terms of regularity of the coeffcients, the minimal assumption is boundedness and measurability, and a unique L2- valued solution is then readily available. We investigate its further properties, such as higher order integrability, boundedness, and continuity. The other class of equations considered here are the ones whose leading operators do not satisfy the strong coercivity condition, but only a degenerate version of it, and therefore are not covered by the classical theory. We derive solvability in Wmp spaces and also discuss their numerical approximation through finite different schemes.

In this thesis a mathematical description and analysis of the Cumulative Prospect Theory is presented. Conditions that ensure well-posedness of the problem are provided, as well as existence results concerning optimal policies for discrete-time incomplete market models and for a family of diffusion market models. A brief outline of how this work is organised follows. In Chapter 2 important results on weak convergence and discrete time finance models are described, these facts form the main background to introduce in Chapter 3 the problem of optimal investment under the CPT theorem in a discrete time setting. We describe our model, present some assumptions and main results are derived. The second part of this work comprises the description of the martingale problem formulation of diffusion processes in Chapter 4. A key result on the limits and topological properties of the set of laws of a class of Itô processes is described in Chapter 5. Finally, we introduce a factor model that includes a class of stochastic volatility models, possibly with path-depending coefficients. Under this model, the problem of optimal investment with a behavioural investor is analysed and our main results on well-posedness and existence of optimal strategies are described under the framework of weak solutions. Further research and challenges when applying the techniques developed in this work are described.

We investigate an explicit tamed Euler scheme of stochastic differential equation with random coefficients driven by Lévy noise, which has super-linear drift coefficient. The strong convergence property of the tamed Euler scheme is proved when drift coefficient satisfies one-sided local Lipschitz condition whereas diffusion and jump coefficients satisfy local Lipschitz conditions. A rate of convergence for the tamed Euler scheme is recovered when local Lipschitz conditions are replaced by global Lipschitz conditions and drift satisfies polynomial Lipschitz condition. These findings are consistent with those of the classical Euler scheme. New methodologies are developed to overcome challenges arising due to the jumps and the randomness of the coefficients. Moreover, as an application of these findings, a tamed Euler scheme is proposed for the stochastic delay differential equation driven by Lévy noise with drift coefficient that grows super-linearly in both delay and non-delay variables. The strong convergence property of the tamed Euler scheme for such SDDE driven by Lévy noise is studied and rate of convergence is shown to be consistent with that of the classical Euler scheme. Finally, an explicit tamed Milstein scheme with rate of convergence arbitrarily close to one is developed to approximate the stochastic differential equation driven by Lévy noise (without random coefficients) that has super-linearly growing drift coefficient.

In this thesis, we study the existence, uniqueness, and regularity of systems of degenerate linear stochastic integro-differential equations (SIDEs) of parabolic type with adapted coefficients in the whole space. We also investigate explicit and implicit finite difference schemes for SIDEs with non-degenerate diffusion. The class of equations we consider arise in non-linear filtering of semimartingales with jumps. In Chapter 2, we derive moment estimates and a strong limit theorem for space inverses of stochastic flows generated by Lévy driven stochastic differential equations (SDEs) with adapted coefficients in weighted Hölder norms using the Sobolev embedding theorem and the change of variable formula. As an application of some basic properties of flows of Weiner driven SDEs, we prove the existence and uniqueness of classical solutions of linear parabolic second order stochastic partial differential equations (SPDEs) by partitioning the time interval and passing to the limit. The methods we use allow us to improve on previously known results in the continuous case and to derive new ones in the jump case. Chapter 3 is dedicated to the proof of existence and uniqueness of classical solutions of degenerate SIDEs using the method of stochastic characteristics. More precisely, we use Feynman-Kac transformations, conditioning, and the interlacing of space inverses of stochastic flows generated by SDEs with jumps to construct solutions. In Chapter 4, we prove the existence and uniqueness of solutions of degenerate linear stochastic evolution equations driven by jump processes in a Hilbert scale using the variational framework of stochastic evolution equations and the method of vanishing viscosity. As an application, we establish the existence and uniqueness of solutions of degenerate linear stochastic integro-differential equations in the L2-Sobolev scale. Finite difference schemes for non-degenerate SIDEs are considered in Chapter 5. Specifically, we study the rate of convergence of an explicit and an implicit-explicit finite difference scheme for linear SIDEs and show that the rate is of order one in space and order one-half in time.

In this work we present some new results concerning stochastic partial differential and integro-differential equations (SPDEs and SPIDEs) that appear in non-linear filtering. We prove existence and uniqueness of solutions of SPIDEs, we give a comparison principle and we suggest an approximation scheme for the non-local integral operators. Regarding SPDEs, we use techniques motivated by the work of De Giorgi, Nash, and Moser, in order to derive global and local supremum estimates, and a weak Harnack inequality.

Our aim is to examine the problem of optimal asset allocation for investors exhibiting a behaviour in the face of uncertainty which is not consistent with the usual axioms of Expected Utility Theory. This thesis is divided into two main parts. In the first one, comprising Chapter II, we consider an arbitrage-free discrete-time financial model and an investor whose risk preferences are represented by a possibly nonconcave utility function (defined on the non-negative half-line only). Under straightforward conditions, we establish the existence of an optimal portfolio. As for Chapter III, it consists of the study of the optimal investment problem within a continuous-time and (essentially) complete market framework, where asset prices are modelled by semi-martingales. We deal with an investor who behaves in accordance with Kahneman and Tversky's Cumulative Prospect Theory, and we begin by analysing the well-posedness of the optimisation problem. In the case where the investor's utility function is not bounded above, we derive necessary conditions for well-posedness, which are related only to the behaviour of the distortion functions near the origin and to that of the utility function as wealth becomes arbitrarily large (both positive and negative). Next, we focus on an investor whose utility is bounded above. The problem's wellposedness is trivial, and a necessary condition for the existence of an optimal trading strategy is obtained. This condition requires that the investor's probability distortion function on losses does not tend to zero faster than a given rate, which is determined by the utility function. Provided that certain additional assumptions are satisfied, we show that this condition is indeed the borderline for attainability, in the sense that, for slower convergence of the distortion function, there does exist an optimal portfolio. Finally, we turn to the case of an investor with a piecewise power-like utility function and with power-like distortion functions. Easily verifiable necessary conditions for wellposedness are found to be sufficient as well, and the existence of an optimal strategy is demonstrated.

Molecular dynamics (MD) computations aim to simulate materials at the atomic level by approximating molecular interactions classically, relying on the Born-Oppenheimer approximation and semi-empirical potential energy functions as an alternative to solving the difficult time-dependent Schrodinger equation. An approximate solution is obtained by discretization in time, with an appropriate algorithm used to advance the state of the system between successive timesteps. Modern MD simulations simulate complex systems with as many as a trillion individual atoms in three spatial dimensions. Many applications use MD to compute ensemble averages of molecular systems at constant temperature. Langevin dynamics approximates the effects of weakly coupling an external energy reservoir to a system of interest, by adding the stochastic Ornstein-Uhlenbeck process to the system momenta, where the resulting trajectories are ergodic with respect to the canonical (Boltzmann-Gibbs) distribution. By solving the resulting stochastic differential equations (SDEs), we can compute trajectories that sample the accessible states of a system at a constant temperature by evolving the dynamics in time. The complexity of the classical potential energy function requires the use of efficient discretization schemes to evolve the dynamics. In this thesis we provide a systematic evaluation of splitting-based methods for the integration of Langevin dynamics. We focus on the weak properties of methods for confiurational sampling in MD, given as the accuracy of averages computed via numerical discretization. Our emphasis is on the application of discretization algorithms to high performance computing (HPC) simulations of a wide variety of phenomena, where configurational sampling is the goal. Our first contribution is to give a framework for the analysis of stochastic splitting methods in the spirit of backward error analysis, which provides, in certain cases, explicit formulae required to correct the errors in observed averages. A second contribution of this thesis is the investigation of the performance of schemes in the overdamped limit of Langevin dynamics (Brownian or Smoluchowski dynamics), showing the inconsistency of some numerical schemes in this limit. A new method is given that is second-order accurate (in law) but requires only one force evaluation per timestep. Finally we compare the performance of our derived schemes against those in common use in MD codes, by comparing the observed errors introduced by each algorithm when sampling a solvated alanine dipeptide molecule, based on our implementation of the schemes in state-of-the-art molecular simulation software. One scheme is found to give exceptional results for the computed averages of functions purely of position.

First we consider implicit finite difference schemes on uniform grids in time and space for second order linear stochastic partial differential equations of parabolic type. Under sufficient regularity conditions, we prove the existence of an appropriate asymptotic expansion in powers of the the spatial mesh and hence we apply Richardson's method to accelerate the convergence with respect to the spatial approximation to an arbitrarily high order. Then we extend these results to equations where the parabolicity condition is allowed to degenerate. Finally, we consider implicit finite difference approximations for deterministic linear second order partial differential equations of parabolic type and give sufficient conditions under which the approximations in space and time can be simultaneously accelerated to an arbitrarily high order.

This thesis presents two dimensional discrete dynamical system, the extended standard family of maps, which approximates homoclinic bifurcations of continuous dissipative systems. The main subject of study is the problem of separatrix splitting which was first discovered by Poincaré in the context of the n-body problem. Separatrix splitting leads to chaotic behaviour of the system on exponentially small region in parameter space. To estimate the size of the region the dissipative map is extended to complex variables and approximated by differential equation on a specific domain. This approach was proposed by Lazutkin to study separatrix splitting for Chirikov’s standard map. Furthermore the complex nearly periodic function is used to estimate the width of the exponentially small region where chaos prevails and the map is related to the semistandard map. Numerical computations require solving complex differential equation and provide the constants involved in the asymptotic formula for the size of the region. Another problem studied in this thesis is the prevalence of resonance for the dissipative standard map on a specific invariant set, which for one dimensional map corresponds to a circle. The regions in parameter space where periodic behaviour occurs on the invariant set is known as Arnold tongues. The width of Arnold tongue is studied and numerical results obtained by iterating the map and solving differential equation are related to the semistandard map.

The Black-Scholes model and corresponding option pricing formula has led to a wide and extensive industry, used by financial institutions and investors to speculate on market trends or to control their level of risk from other investments. From the formation of the Chicago Board Options Exchange in 1973, the nature of options contracts available today has grown dramatically from the single-date contracts considered by Black and Scholes (1973) to a wider and more exotic range of derivatives. These include American options, which can be exercised at any time up to maturity, as well as options based on the weighted sums of assets, such as the Asian and basket options which we consider. Moreover, the underlying models considered have also grown in number and in this work we are primarily motivated by the increasing interest in past-dependent asset pricing models, shown in recent years by market practitioners and prominent authors. These models provide a natural framework that considers past history and behaviour, as well as present information, in the determination of the future evolution of an underlying process. In our studies, we explore option pricing techniques for arithmetic Asian and basket options under a Stochastic Delay Differential Equation (SDDE) approach. We obtain explicit closed-form expressions for a number of lower and upper bounds before giving a practical, numerical analysis of our result. In addition, we also consider the properties of the approximate numerical integration methods used and state the conditions for which numerical stability and convergence can be achieved.

The evaluation of the product of two matrices can be very computationally expensive. The multiplication of two n×n matrices, using the “default” algorithm can take O(n3) field operations in the underlying field k. It is therefore desirable to find algorithms to reduce the “cost” of multiplying two matrices together. If multiplication of two n × n matrices can be obtained in O(nα) operations, the least upper bound for α is called the exponent of matrix multiplication and is denoted by ω. A bound for ω < 3 was found in 1968 by Strassen in his algorithm. He found that multiplication of two 2 × 2 matrices could be obtained in 7 multiplications in the underlying field k, as opposed to the 8 required to do the same multiplication previously. Using recursion, we are able to show that ω ≤ log2 7 < 2.8074, which is better than the value of 3 we had previously. In chapter 1, we look at various techniques that have been found for reducing ω. These include Pan’s Trilinear Aggregation, Bini’s Border Rank and Sch¨onhage’s Asymptotic Sum inequality. In chapter 2, we look in detail at the current best estimate of ω found by Coppersmith and Winograd. We also propose a different method of evaluating the “value” of trilinear forms. Chapters 3 and 4 build on the work of Coppersmith and Winograd and examine how cubing and raising to the fourth power of Coppersmith and Winograd’s “complicated” algorithm affect the value of ω, if at all. Finally, in chapter 5, we look at the Group-Theoretic context proposed by Cohn and Umans, and see how we can derive some of Coppersmith and Winograd’s values using this method, as well as showing how working in this context can perhaps be more conducive to showing ω = 2.

This work is concerned with an evolution inclusion of a form, in a triple of spaces \V -> H -> V*", where U is a continuous non-decreasing process, M is a locally square-integrable martingale and the operators A (multi-valued) and B satisfy some monotonicity condition, a coercivity condition and a condition on growth in u. An existence and uniqueness theorem is proved for the solutions, using semi-implicit time-discretization schemes. Examples include evolution equations and inclusions driven by square integrable Levy martingales.

We study numerical approximations for the payoff function of the stochastic optimal stopping and control problem. It is known that the payoff function of the optimal stopping and control problem corresponds to the solution of a normalized Bellman PDE. The principal aim of this thesis is to study the rate at which finite difference approximations, derived from the normalized Bellman PDE, converge to the payoff function of the optimal stopping and control problem. We do this by extending results of N.V. Krylov from the Bellman equation to the normalized Bellman equation. To our best knowledge, until recently, no results about the rate of convergence of finite difference approximations to Bellman equations have been known. A major breakthrough has been made by N. V. Krylov. He proved rate of rate of convergence of tau 1/4 + h 1/2 where tau and h are the step sizes in time and space respectively. We will use the known idea of randomized stopping to give a direct proof showing that optimal stopping and control problems can be rewritten as pure optimal control problems by introducing a new control parameter and by allowing the reward and discounting functions to be unbounded in the control parameter. We extend important results of N. V. Krylov on the numerical solutions to the Bellman equations to the normalized Bellman equations associated with the optimal stopping of controlled diffusion processes. We obtain the same rate of convergence of tau1/4 + h1/2. This rate of convergence holds for finite difference schemes defined on a grid on the whole space [0, T]×Rd i.e. on a grid with infinitely many elements. This leads to the study of localization error, which arises when restricting the finite difference approximations to a cylindrical domain. As an application of our results, we consider an optimal stopping problem from mathematical finance: the pricing of American put option on multiple assets. We prove the rate of convergence of tau1/4 + h1/2 for the finite difference approximations.

In the work we estimate the rate of convergence of the Wong-Zakai type of approximations for SDEs and SPDEs. Two cases are studied: SDEs in finite dimensional settings and evolution stochastic systems (SDEs in the infinite dimensional case). The latter result is applied to the second order SPDEs of parabolic type and the filtering problem. Roughly, the result is the following. Let Wn be a sequence of continuous stochastic processes of finite variation on an interval [0, T]. Assume that for some a > 0 the processes Wn converge almost surely in the supremum norm in [0, T] to W with the rate n-k for each k < a. Then the solutions Un of the differential equations with Wn converge almost surely in the supremum norm in [0, T] to the solution u of the "Stratonovich" SDE with W with the same rate of convergence, n-k for each k < a, in the case of SDEs and with the rate of convergence n-k/2 for each k < a, in the case of evolution systems and SPDEs. In the final chapter we verify that the two most common approximations of the Wiener process, smoothing and polygonal approximation, satisfy the assumptions made in the previous chapters.