A survey of evolutionary algorithms for decision-tree induction

This paper presents a survey of evolutionary algorithms that are designed for decision-tree induction. In this context, most of the paper focuses on approaches that evolve decision trees as an alternate heuristics to the traditional top-down divide-and-conquer approach. Additionally, we present some alternative methods that make use of evolutionary algorithms to improve particular components of decision-tree classifiers. The paper's original contributions are the following. First, it provides an up-to-date overview that is fully focused on evolutionary algorithms and decision trees and does not concentrate on any specific evolutionary approach. Second, it provides a taxonomy, which addresses works that evolve decision trees and works that design decision-tree components by the use of evolutionary algorithms. Finally, a number of references are provided that describe applications of evolutionary algorithms for decision-tree induction in different domains. At the end of this paper, we address some important issues and open questions that can be the subject of future research.

Staining of the Internal Limiting Membrane with the Use of Heavy Brilliant Blue G

Background: Brilliant blue G (BBG) is frequently used in chromovitrectomy to facilitate internal limiting membrane (ILM) peeling. A study was initiated to evaluate if heavy BBG is safe and effective in staining the ILM. Methods: We studied 30 eyes, 23 with idiopathic macular holes and 7 of patients with diabetic macular edema. Removal of the ILMs was assisted by heavy BBG staining. In cases with histopathological correlation the ILMs were evaluated with hematoxylin and eosin, Masson's trichrome, periodic acid-Schiff and glial fibrillary acidic protein staining. In addition, immunohistochemistry was also performed using specific antibodies for vimentin, neuron-specific enolase, factor VIII and CD68. Using the Image-Pro Plus software of Media Cybernetics Co. we found an average thickness in ILMs. Results: Of the ILM specimens sent, 19/30(63.33%) could not be processed properly because of the limited sample material, recognizing only fragments of dispersed fibrillar material. In macular hole ILMs we found an average thickness of 1.3 +/- 0.65 mu m, and in diabetic macular edema ILMs an average thickness of 6.2 +/- 1.4 mu m. Conclusions: In heavy BBG-assisted ILM peeling we observed no intraoperative or postoperative complications after a mean follow-up of 12 months. Heavy BBG could be an effective and safe vehicle for staining the ILM. Copyright (C) 2012 S. Karger AG, Basel

Uma arqueologia de diagramas cibernéticos

This paper investigates the use of explicit structures of information in architectural design. Particularly, it approaches the use of diagrams related to cybernetics and information theory in experimental practices in the 1960’s and 1970’s. It analyses the diagram of cybernetic control proposed by the cybernetician Gordon Pask for the Fun Palace, the diagrams produced by the utopian architect Yona Friedman in the conceptual description of the Flatwriter program and Christopher Alexander’s diagrams and his theories of Synthesis of Form and Pattern Language. Finally it establishes a brief parallel between current domestication and use of dataflow programming with the cybernetic diagrams, highlighting differences in their complexity approach.

FreeConj-TIP

[EN] FreeConj-TIP es programa desarrollado en el lenguaje de programación C++, bajo licencia GNU General Public, que realiza la conjugación de verbos del español. FreeConj-TIP resuelve la conjugación aplicando reglas elementales de la ortografía española y de conjugación para verbos con una cierta irregularidad en un 97,42% de un total de 14613 verbos, y el 2,57%, restante, correspondiente a verbos muy irregulares, se resuelve en base a datos. El objetivo principal del proyecto ha sido el de construir un conjugador verbal con una implementación sencilla, pero sin que por ello existan verbos que no puedan ser conjugados de acuerdo a todos los modelos de conjugación que les resulten aplicables.

Tor network forensics and hidden service deanonymization

The cybernetics revolution of the last years improved a lot our lives, having an immediate access to services and a huge amount of information over the Internet. Nowadays the user is increasingly asked to insert his sensitive information on the Internet, leaving its traces everywhere. But there are some categories of people that cannot risk to reveal their identities on the Internet. Even if born to protect U.S. intelligence communications online, nowadays Tor is the most famous low-latency network, that guarantees both anonymity and privacy of its users. The aim of this thesis project is to well understand how the Tor protocol works, not only studying its theory, but also implementing those concepts in practice, having a particular attention for security topics. In order to run a Tor private network, that emulates the real one, a virtual testing environment has been configured. This behavior allows to conduct experiments without putting at risk anonymity and privacy of real users. We used a Tor patch, that stores TLS and circuit keys, to be given as inputs to a Tor dissector for Wireshark, in order to obtain decrypted and decoded traffic. Observing clear traffic allowed us to well check the protocol outline and to have a proof of the format of each cell. Besides, these tools allowed to identify a traffic pattern, used to conduct a traffic correlation attack to passively deanonymize hidden service clients. The attacker, controlling two nodes of the Tor network, is able to link a request for a given hidden server to the client who did it, deanonymizing him. The robustness of the traffic pattern and the statistics, such as the true positive rate, and the false positive rate, of the attack are object of a potential future work.

Learning flexible sensori-motor mappings in a complex network

Given the complex structure of the brain, how can synaptic plasticity explain the learning and forgetting of associations when these are continuously changing? We address this question by studying different reinforcement learning rules in a multilayer network in order to reproduce monkey behavior in a visuomotor association task. Our model can only reproduce the learning performance of the monkey if the synaptic modifications depend on the pre- and postsynaptic activity, and if the intrinsic level of stochasticity is low. This favored learning rule is based on reward modulated Hebbian synaptic plasticity and shows the interesting feature that the learning performance does not substantially degrade when adding layers to the network, even for a complex problem.

Neural Networks as Cybernetic Systems - Part I

Neural Networks as Cybernetic Systems is a textbox that combines classical systems theory with artificial neural network technology.

Neural Networks as Cybernetic Systems - Part II

Neural Networks as Cybernetic Systems is a textbox that combines classical systems theory with artificial neural network technology.

tkCybernetics

tkCybernetics is a simulation software that can be used to create and evaluate simple cybernetic circuits as a combination of lowpass and highpass filters and several nonlinear elements. tkCybernetics has been developed for and successfully applied to introductory courses on biological cybernetics at Bielefeld University. Therefore, it is recommended for students who start learning about cybernetics. Students can quickly grasp the concepts used in the graphical user interface and are able to create their own circuits after a short introduction to the program. Its usefulness for advanced students and researchers is limited to quickly testing circuits. This article provides detailed tutorial information on how to install and work with tkCybernetics for self-learning and for higher education.

Neural Networks as Cybernetic Systems

This book on biological cybernetics combines system theory and artificial neural networks. Following a 'holistic' approach, the book examines the role of simulation in biology. Mainly addressed to students of Biology, the book tries to avoid the use of mathematical formula as far as possible. Exercises can be performed with a related software tool (tkCybernetics) for part 1 as well as with educational simulation on neural networks for part two (in prepration.)

Neural Networks as Cybernetic Systems - 3rd and revised edition

This third edition essentially compares with the 2nd one, but has been improved by correction of errors and by a rearrangement and minor expansion of the sections referring to recurrent networks. These changes hopefully allow for an easier comprehension of the essential aspects of this important domain that has received growing attention during the last years.