Different methods for Blind Source Separation (BSS) have been recently proposed. Most of these methods are suitable for separating either a mixture of sub-Gaussian source or a mixture of super-Gaussian sources. In this paper, a unified statistical approach for separating the mixture of sub-Gaussian and super-Gaussian source is proposed. Source separation techniques use an objective function to be optimized. The optimization process requires probability density function to be expressed in the terms of the random variable. Two different density models have been used for representing sub-Gaussian and super-Gaussian sources. Optimization of the objective function yields different nonlinear functions. Kurtosis has been ušed as measure of Gaussianity of a source. Depending upon the sign of kurtosis one of the nonlinearities is ušed in the proposed algorithm. Simulations with artificiaily generated as well as audio signals demonstrate effectiveness of the proposed approach.
A short approach to the Kurzweil-Henstock integral is outlined, based on approximating a real function on a compact interval by suitable step-functions, and using filterbase convergence to define the integral. The properties of the integral are then easy to establish.
The contribution focuses on the design of a control algorithm aimed at the operative control of runoff water from a reservoir during flood situations. Management is based on the stochastically specified forecast of water inflow into the reservoir. From a mathematical perspective, the solved task presents the control of a dynamic system whose predicted hydrological input (water inflow) is characterised by significant uncertainty. The algorithm uses a combination of simulation model data, in which the position of the bottom outlets is sought via nonlinear optimisation methods, and artificial intelligence methods (adaptation and fuzzy model). The task is written in the technical computing language MATLAB using the Fuzzy Logic Toolbox.
In this paper, we consider a distributed stochastic computation of AXB=C with local set constraints over an multi-agent system, where each agent over the network only knows a few rows or columns of matrixes. Through formulating an equivalent distributed optimization problem for seeking least-squares solutions of AXB=C, we propose a distributed stochastic mirror-descent algorithm for solving the equivalent distributed problem. Then, we provide the sublinear convergence of the proposed algorithm. Moreover, a numerical example is also given to illustrate the effectiveness of the proposed algorithm.
Maintaining liquid asset portfolios involves a high carry cost and is mandatory by law for most financial institutions. Taking this into account a financial institution's aim is to manage a liquid asset portfolio in an "optimal" way, such that it keeps the minimum required liquid assets to comply with regulations. In this paper we propose a multi-stage dynamic stochastic programming model for liquid asset portfolio management. The model allows for portfolio rebalancing decisions over a multi-period horizon, as well as for flexible risk management decisions, such as reinvesting coupons, at intermediate time steps. We show how our problem closely relates to insurance products with guarantees and utilize this in the formulation. We will discuss our formulation and implementation of a multi-stage stochastic programming model that minimizes the down-side risk of these portfolios. The model is back-tested on real market data over a period of two years
In a Discounted Markov Decision Process (DMDP) with finite action sets the Value Iteration Algorithm, under suitable conditions, leads to an optimal policy in a finite number of steps. Determining an upper bound on the necessary number of steps till gaining convergence is an issue of great theoretical and practical interest as it would provide a computationally feasible stopping rule for value iteration as an algorithm for finding an optimal policy. In this paper we find such a bound depending only on structural properties of the Markov Decision Process, under mild standard conditions and an additional "individuality" condition, which is of interest in its own. It should be mentioned that other authors find such kind of constants using non-structural information, i.e., information not immediately apparent from the Decision Process itself. The DMDP is required to fulfill an ergodicity condition and the corresponding ergodicity index plays a critical role in the upper bound.
In this paper we obtain a strong invariance principle for negatively associated random fields, under the assumptions that the field has a finite $(2+\delta )$th moment and the covariance coefficient $u(n)$ exponentially decreases to $0$. The main tools are the Berkes-Morrow multi-parameter blocking technique and the Csörgő-Révész quantile transform method.
Two interacting active regions of different ages have been studied over two days using white light and Hα observations as well as magnetograma. Different series of homologous flares and the formation of a filament in tbe region were analysed and an interpretation of these events is proposed.
Let F be a class of entire functions represented by Dirichlet series with complex frequencies ∑ ake hλ k ,zi for which (|λ k |/e)|λ k | k!|ak| is bounded. Then F is proved to be a commutative Banach algebra with identity and it fails to become a division algebra. F is also proved to be a total set. Conditions for the existence of inverse, topological zero divisor and continuous linear functional for any element belonging to F have also been established.
In this paper, we study decentralized H∞ feedback control systems with quantized signals in local input-output (control) channels. We first assume that a decentralized output feedback controller has been designed for a multi-channel continuous-time system so that the closed-loop system is Hurwitz stable and a desired H∞ disturbance attenuation level is achieved. However, since the local measurement outputs are quantized by a general quantizer before they are passed to the controller, the system's performance is not guaranteed. For this reason, we propose a local-output-dependent strategy for updating the quantizers' parameters, so that the closed-loop system is asymptotically stable and achieves the same H∞ disturbance attenuation level. We also extend the discussion and the result to the case of multi-channel discrete-time H∞ feedback control systems.