This research studies how to efficiently predict optimal active constraints of an inequality constrained optimization problem, in the context of Interior Point Methods (IPMs). We propose a framework based on shifting/perturbing the inequality constraints of the problem. Despite being a class of powerful tools for solving Linear Programming (LP) problems, IPMs are well-known to encounter difficulties with active-set prediction due essentially to their construction. When applied to an inequality constrained optimization problem, IPMs generate iterates that belong to the interior of the set determined by the constraints, thus avoiding/ignoring the combinatorial aspect of the solution. This comes at the cost of difficulty in predicting the optimal active constraints that would enable termination, as well as increasing ill-conditioning of the solution process. We show that, existing techniques for active-set prediction, however, suffer from difficulties in making an accurate prediction at the early stage of the iterative process of IPMs; when these techniques are ready to yield an accurate prediction towards the end of a run, as the iterates approach the solution set, the IPMs have to solve increasingly ill-conditioned and hence difficult, subproblems. To address this challenging question, we propose the use of controlled perturbations. Namely, in the context of LP problems, we consider perturbing the inequality constraints (by a small amount) so as to enlarge the feasible set. We show that if the perturbations are chosen judiciously, the solution of the original problem lies on or close to the central path of the perturbed problem. We solve the resulting perturbed problem(s) using a path-following IPM while predicting on the way the active set of the original LP problem; we find that our approach is able to accurately predict the optimal active set of the original problem before the duality gap for the perturbed problem gets too small. Furthermore, depending on problem conditioning, this prediction can happen sooner than predicting the active set for the perturbed problem or for the original one if no perturbations are used. Proof-of-concept algorithms are presented and encouraging preliminary numerical experience is also reported when comparing activity prediction for the perturbed and unperturbed problem formulations. We also extend the idea of using controlled perturbations to enhance the capabilities of optimal active-set prediction for IPMs for convex Quadratic Programming (QP) problems. QP problems share many properties of LP, and based on these properties, some results require more care; furthermore, encouraging preliminary numerical experience is also presented for the QP case.

Prof. Marcel Wild completed his PhD with Zurick University and this is a copy of the original works / The original works can be found at http://www.hbz.uzh.ch/ / ABSTRACT: A linear representation of a modular lattice L is a homomorphism from L into the lattice Sub(V) of all subspaces of a vector space V. The representation theory of lattices was initiated by the Darmstadt school (Wille, Herrmann, Poguntke, et al), to large extent triggered by the linear representations of posets (Gabriel, Gelfand-Ponomarev, Nazarova, Roiter, Brenner, et al). Even though posets are more general than lattices, none of the two theories encompasses the other. In my thesis a natural type of finite lattice is identified, i.e. triangle lattices, and their linear representation theory is advanced. All of this was triggered by a more intricate setting of quadratic spaces (as opposed to mere vector spaces) and the question of how Witt’s Theorem on the congruence of finite-dimensional quadratic spaces lifts to spaces of uncountable dimensions. That issue is dealt with in the second half of the thesis.

This report reviews the 2007 financial crisis and the actions of the Federal Reserve. The Full Employment Act of 1946 and the "Humphrey-Hawkins" Act guides the Fed's actions. These two laws outline the long-term goals of the monetary policy framework the Fed uses; however, the framework lacks principles for achieving the mandated long term goals such as reliable, complete data. This report looks at the use of model-based forecasting and gives recommendations for principles which will strengthen the preexisting monetary framework. / text

Techniques for facial expression generation are employed in several applications in computer graphics as well as in the processing of image and video sequences containing faces. Video coding standards such as MPEG-4 support facial expression animation. There are a number of facial expression representations that are application dependent or facial animation standard dependent and most of them require a lot of computational effort. We have developed a completely novel and effective method for representing the primary facial expressions using a model-independent set of deformation parameters (derived using rubber-sheet transformations), which can be easily applied to transform facial feature points. The developed mathematical model captures the necessary non-linear characteristics of deformations of facial muscle regions; producing well-recognizable expressions on images, sketches, and three dimensional models of faces. To show the effectiveness of the method, we developed a variety of novel applications such as facial expression recognition, expression mapping, facial animation and caricature generation.

The introduction of analytical techniques along the steps of the development of a space launcher will allow significant reductions in terms of costs and manpower, and will enable, by a more systematical way of tuning and assessing control laws, to get flyable designs much faster. In this scope, IQC based tools already present promising result and show that they may be the most appropriate ones for the robustness analysis of large complex systems. They account for the system structure and allow dealing specifically with each subsystems, it means that we can improve the representation contained in the multipliers easily and reuse the set up to assess the improvements. The flexibility of the method is a huge advantage. We experienced it during two phases. The first was dedicated to the analysis of the three-degree-of-freedom uncertain nonlinear equation of motion of a rigid body. Secondly, we studied the influence of the pulse-width modulator behavior of the attitude control system on the launcher stability. IQC-based stability analysis allowed defining estimations of the stability domain with respect to uncertainties and system parameters. Moreover, the results obtained with IQC can go way beyond stability analysis with performance analysis with description of the particular performance criteria of the field with appropriate multipliers. Later on controller synthesis and merging of IQC method with worst-case search algorithms could extend greatly the frame of use of this analytical tool and give it the influence it deserves.

No description available.

While stochastic optimal control, together with associate formulations like Reinforcement Learning, provides a formal approach to, amongst other, motor control, it remains computationally challenging for most practical problems. This thesis is concerned with the study of relations between stochastic optimal control and probabilistic inference. Such dualities { exempli ed by the classical Kalman Duality between the Linear-Quadratic-Gaussian control problem and the filtering problem in Linear-Gaussian dynamical systems { make it possible to exploit advances made within the separate fields. In this context, the emphasis in this work lies with utilisation of approximate inference methods for the control problem. Rather then concentrating on special cases which yield analytical inference problems, we propose a novel interpretation of stochastic optimal control in the general case in terms of minimisation of certain Kullback-Leibler divergences. Although these minimisations remain analytically intractable, we show that natural relaxations of the exact dual lead to new practical approaches. We introduce two particular general iterative methods ψ-Learning, which has global convergence guarantees and provides a unifying perspective on several previously proposed algorithms, and Posterior Policy Iteration, which allows direct application of inference methods. From these, practical algorithms for Reinforcement Learning, based on a Monte Carlo approximation to ψ-Learning, and model based stochastic optimal control, using a variational approximation of posterior policy iteration, are derived. In order to overcome the inherent limitations of parametric variational approximations, we furthermore introduce a new approach for none parametric approximate stochastic optimal control based on a reproducing kernel Hilbert space embedding of the control problem. Finally, we address the general problem of temporal optimisation, i.e., joint optimisation of controls and temporal aspects, e.g., duration, of the task. Specifically, we introduce a formulation of temporal optimisation based on a generalised form of the finite horizon problem. Importantly, we show that the generalised problem has a dual finite horizon problem of the standard form, thus bringing temporal optimisation within the reach of most commonly used algorithms. Throughout, problems from the area of motor control of robotic systems are used to evaluate the proposed methods and demonstrate their practical utility.

The American vaccine pricing market has many actors, making it a complex system to model. Because of this, previous papers have chosen to model only vaccine manufacturers while leaving out the government. However, the government is also an important actor in the market, since it buys over half of vaccines produced. In this work, we aim to introduce the government into vaccine pricing models to better recommend pricing strategies to the Centers for Disease Control and Prevention.

The “infrastructure” of quadratic fields is a body of theory developed by Dan Shanks, Richard Mollin and others, in which they relate “reduced ideals” in the rings and sub-rings of integers in quadratic fields with periodicity in continued fraction expansions of quadratic numbers. In this thesis, we develop cubic analogs for several infrastructure theorems. We work in the field K=Q(), where 3=m for some square-free integer m, not congruent to ±1, modulo 9. First, we generalize the definition of a reduced ideal so that it applies to K, or to any number field. Then we show that K has only finitely many reduced ideals, and provide an algorithm for listing them. Next, we define a sequence based on the number alpha that is periodic and corresponds to the finite set of reduced principal ideals in K. Using this rudimentary infrastructure, we are able to establish results about fundamental units and reduced ideals for some classes of pure cubic fields. We also introduce an application to Diophantine approximation, in which we present a 2-dimensional analog of the Lagrange value of a badly approximable number, and calculate some examples.

The aim of the thesis is to find out students' reasoning of chosen word problem by using quadratic equation. The work focuses on specific errors end problems reported by students and evaluated by Newman's method of Error Causes for Written Mathematical Tasks. The theoretical work was based on analyse of current mathematical textbooks and comparison with several international pedagogical studies and thesis with similar specialization. Furthermore, a detailed description of methodology and my own research are described in practical part of the thesis. Principle of study was to chose group of the students from two different high schools ( one well know grammar school and one business high school in Prague) and record the process of reasoning the word problem by camera. The concluison is dedicated to proper analyses of mistakes and problems during the student's reasonings. Powered by TCPDF (www.tcpdf.org)