Catálogo de publicaciones - libros

Classification of internet packets is a common task for internet routers. More rapid information transformation in communication links requires better processing and classifying algorithms. This paper presents a novel classification algorithm for internet packets profits parallel processing of neural networks. It is based on an extended able design of simple neural network blocks which are fast to learn and fast to respond. The classifier could be updated by re-learning simple neural networks, whenever a rule changes.

Customer call centers are the preferred and prevalent way for many companies to communicate with their customers. The customer call center industry is thus vast and rapidly expanding in terms of both workforce and economic scope. Most major companies have reengineered their communication with customers via one or more call centers, either internally managed or outsourced. Call centers constitutes a set of resources which enable the delivery of services via telephone, email or web portal access. Customer inquiries contains different types of uncertainties regarding the problem description, the recommended system solution and precise cause study. We develop a decision support system for customer call centers using soft computing techniques for automating, maintaining and maximizing the value of the decision process. Fuzzy logic as soft computing technique is a methodology for the representation and manipulation of imprecise and vague information. Bayesian networks are formal graphical languages for the representation and communication of decision scenarios requiring reasoning under uncertainty. We discuss decision support system scenarios under uncertainty using Bayesian networks and fuzzy logic. Real customer requests as support cases contain cause action coherence under uncertainty. We will model these types of uncertainty scenarios in a decision support system selecting the appropriate technique of supporting the decision process.

This paper introduces a new application of the genetic algorithm for online control application. It acts as a model free optimization technique that belongs to the class of reinforcement learning. Its concepts and structure is first investigated and then the ability of this algorithm is highlighted by an application in a real-time control (pole balancing) problem. The simulation results approves the better the merit of the proposed technique.

This paper is the first step of a multi-sensor fusion system for control of dept of desflurane anesthesia. In this study, depth of desflurane anesthesia was examined through cardiovascular-based an adaptive neuro-fuzzy system according to changing in the blood pressure and heart rate taken from the patient. The second step, in the next paper will be based on auditory evoked responses. The system designed for anesthetic agent, desflurane, because it is very popular and among the first choices of anesthesiologist for inhalation anesthesia. Intraoperative awareness resulting from inadequate anesthetic is a rare but serious complication during general anesthesia. In order to prevent possible intraoperative awareness, anesthesiologists usually apply anesthetics at level much above the minimal necessary. Anesthetic overdosing prolongs the recovery period, which may cause severe hemodynamic depression and a life-threatening scenario in critically ill patients. To increase patient safety and comfort is one of the most important potential benefits of the system. The second important aim of the study is to relase the anesthesiologist so that he or she can devote attention to other tasks that can’t yet be adequately automated. Also, to make the optimum in the area of anesthetic agent and to economize by lessening the costs of an operation are included the benefits which are coming with this system.

Therapeutic ultrasound is an emerging field with many medical applications. High intensity focused ultrasound provides the ability to localize the deposition of acoustic energy within the body, which can cause tissue necrosis and hemostasis. The ultrasound applied in therapy is usually ranged from 1MHz to 1000MHz. Even the least vibration of 1MHz would be as keen as a sharp knife to cut off steels, if we reinforce its amplitude. However, the output of the ultrasound used in treating people must be decreased substantially. A specific increase in temperature is necessary to achieve a temperature-mediated therapeutic impact by ultrasound in rehabilitation. On a large scale ultrasound intensity determines the temperature level on the tissue. High intensity causes a marked mechanical peak loading of the tissue. This may even lead to tissue damage. The extreme pressure differences developing as a consequence of exposure to ultrasound may cause cavitations in the tissues. Opinions in the literature on the duration of treatment also vary. The duration of treatment depends on the size of the body area to be treated. Lehmann fixes the maximum duration of treatment at 15 minutes. This refers to a treated area of 75–100 cm^2 which he considers the maximum area that can reasonably be treated. New medical applications have required advances in biomedical equipment design and advances in numerical and experimental studies of the interaction of sound with biological tissues and fluids. In this study a fuzzy logic control system will be explained which was developed in order to obtain optimum ultrasound intensity and determine optimum treatment time during ultrasound therapy (UT). This system also increases patient safety and comfort during UT.