Theory and applications of stochastic processes. An analytical approach (Q1039512): Difference between revisions

This book provides a new insight into stochastic processes for those who are not primarily probabilists and prefer a more physical approach to this theory. In this spirit the main emphasis is not on the mathematical theory but more on its usefulness to (mostly physical) applications. Definitions and new notions are often given as generalizations after motivating examples. With this method also comes a vast amount of examples and exercises throughout all chapters. In the first chapter an introduction into the physical Brownian motion is given beginning with Einstein's movement of a Brownian particle and proceeding with Langevin's equation. It also contains a first introduction into Gaussian processes and the theory of noise. The second chapter supplies the reader with the basics of probability theory, like Markov times and, as an important example that is to be studied much more thoroughly later in this book, the first passage time of a stochastic process. After having seen what a process should be like to be a good model for the physical Brownian motion in the first chapter, this chapter also includes the definition of the mathematical Brownian motion together with two possibilities of its construction. It concludes with important properties, like the non-differentiability, Markov property or martingale property. The second chapter already justified the need for Itô integration and Itô calculus as an alternative to classical differential calculus. The theory of stochastic integration is therefore introduced in the third chapter. From beginning on three different ways to define a stochastic integral (Itô, Stratonovich, backward) are considered and studied side-by-side. Focal points in this discussions are the Itô integral with its martingale property and continuity and the Stratonovich integral with the advantage of satisfying the fundamental theorem of integral calculus and the integration by parts formula. Itô's formula and the Stratonovich chain rule round up this part. Chapter four is about stochastic differential equations, starting with existence and uniqueness results for Itô- and Stratonovich-type equations in the Lipschitz case. Itô's formula can be applied to simplify equations and solutions have the Markov property. Other important results are Kolmogorov's representation formula for the Backward Kolmogorov Equation and the Feynman-Kac-formula. In chapter five the focus of this book goes back to a more applied point and presents approximation possibilities like the Euler simulation scheme and results for different boundary behavior of stochastic systems, i.e. for absorbing, reflecting and partially reflecting boundaries. The sixth chapter studies the first passage time of diffusions with parameters that are time-independent. This way it is possible to use a different approach and get densities for the first passage time as well as explicit calculations for the one-dimensional case. In chapter seven and eight diffusion processes are used to approximate Markov (jump) processes and Langevin's equation. A strong emphasis lies on the boundary behavior of these approximations. The part of Langevin's equation is divided into the cases of overdamped and underdamped Langevin's equation. As some methods like the Kramers-Moyal expansion is not applicable for studying large time scales that are nevertheless necessary to observe extreme events in diffusion processes, chapter nine applies large deviations theory to this problem. Chapter ten includes results for the exit time from a domain for different types of equations and boundaries. The book terminates with the question of stochastic stability in the eleventh chapter. Examples present the case of destabilizing effect of noise for stable deterministic systems as well as the contrary case.