Show simple item record

dc.contributorPabón Burbano, María Constanza
dc.creatorBastidas Caicedo, Harvey Demian
dc.date2018-01
dc.date.accessioned2018-10-24T15:09:00Z
dc.date.available2018-10-24T15:09:00Z
dc.identifier.citationBastidas Caicedo, H. D. (2018, enero ) Computación Evolutiva Descentralizada de Modelo Híbrido usando Blockchain y Prueba de Trabajo de Optimización. Pontificia Universidad Javeriana, Cali.spa
dc.identifier.urihttp://hdl.handle.net/11522/10811
dc.descriptionLas técnicas de Computación Evolutiva (EC) como algoritmos genéticos, neuroevolución o swarm intelligence son métodos de optimización de parámetros de modelos matemáticos que se caracterizan por el uso de una población de soluciones candidatas que evolucionan en un espacio de búsqueda de una forma inspirada en los principios de la evolución biológica como la competencia, la selección o la reproducción. Existen varios modelos arquitecturales para implementar técnicas de EC en arquitecturas de procesamiento distribuido (dEC), los modelos con mejor tolerancia a fallas y menor costo comunicacional son los modelos híbridos basados en el modelo de Islas. Múltiples modelos para dEC se han implementado en frameworks o plataformas de software, pero las implementaciones encontradas tienen desventajas como que tienen baja tolerancia a fallas o les faltan mecanismos de trazabilidad que podrían ser deseables o necesarios para algunas aplicaciones. Para contrarrestar las desventajas mencionadas, el blockchain y la prueba de trabajo criptográfica (CPoW) son tecnologías para almacenar datos de trazabilidad de eventos en redes descentralizadas, pero con el requerimiento de una capacidad computacional adicional para la generación de una CPoW. En este documento se propone el uso de un blockchain con una prueba de trabajo de optimización (OPoW) para implementar un servicio de timestamping en una red descentralizada para optimización con dEC de modelos híbridos, usando para optimización la capacidad computacional que es usada para la generación de una prueba de trabajo criptográfica en otras redes basadas en blockchain como Bitcoin. El sacrificio de usar una prueba de trabajo útil es que la OPoW no es una función del contenido del bloque, sino solo del estado de optimización. Para la validación empírica de la OPoW, este documento describe el diseño de una plataforma de software descentralizada para implementar Algoritmos Evolutivos Distribuidos (dEA) utilizando el modelo de isla y los modelos híbridos. La plataforma propuesta explora el uso de un blockchain y una prueba de trabajo de optimización para almacenar un registro de operaciones para la trazabilidad y la sincronización de los estados de optimización de los nodos participantes en los procesos de dEC. La plataforma propuesta fue implementada y una aplicación para aprendizaje por refuerzo usando dEC en el dominio de automatización de comercio de divisas se usó para realizar experimentos para validar la escalabilidad, tolerancia a fallos y rechazo de resultados inválidos que provee el uso de OPoW.spa
dc.description.abstractThe Evolutionary Computation (EC) techniques such as genetic algorithms, neuroevolution or swarm intelligence are optimization methods characterized by using a population of candidate solutions that evolve in a search space in a way inspired by biological evolution principles like competition, selection or reproduction. There are several architectural models for implementing EC techniques in distributed processing architectures (dEC), the models with better fault tolerance and lower communicational cost are the hybrid models based on the so-called island model. Multiple models for dEC have been implemented in frameworks or software platforms, but the existing implementations are either programming language-specific, lack fault-tolerance or lack traceability features that could be desirable or required for some applications. For counteracting the mentioned disadvantages, the blockchain and the hash-based cryptographic proof-of-work (CPoW) are technologies that allow the storage of data for the traceability of events in a decentralized network, but with an additional computational capacity requirement for the generation of a CPoW. This document proposes the use of a blockchain with an optimization proof-of-work (OPoW) to implement a timestamping service in a decentralized network for dEC with hybrid models, using for optimization the computational capacity that is used for hash-based proof-of-work generation in other blockchain based networks like Bitcoin. The tradeoff of a useful proof of work is that the OPoW is not a function of the block contents but only of the optimization state. For the empirical validation of the proposed proof of work, this document describes the design of a decentralized software platform for implementing distributed Evolutionary Algorithms (dEA) using the island model and hybrid models. The proposed platform explores the use of a blockchain and an optimization proof-of-work to store a log of operations for traceability and synchronization of optimization states of the participating nodes in dEC processes. The proposed platform was implemented and an application for reinforcement learning using dEC in the domain of foreign exchange trading automation was used to perform experiments to validate the scalability, fault-tolerance and invalid result rejection capabilities provided using OPoW.spa
dc.formatapplication/pdfspa
dc.format.extent72 páginasspa
dc.languagespaspa
dc.publisherPontificia Universidad Javerianaspa
dc.rightsEl o los autores otorgan licencia de uso parcial de la obra a favor de la Pontificia Universidad Javeriana Seccional Cali, teniendo en cuenta que en cualquier caso, la finalidad perseguida siempre será facilitar, difundir y promover el aprendizaje, la enseñanza y la investigación. Con la licencia el o los autores autorizan a la Pontificia Universidad Javeriana Seccional Cali: la publicación en formato o soporte material, de acuerdo con las condiciones internas que la Universidad ha establecido para estos efectos. La edición o cualquier otra forma de reproducción, incluyendo la posibilidad de trasladarla al sistema o entorno digital. La inclusión en cualquier otro formato o soporte como multimedia, colecciones, recopilaciones o, en general, servir de base para cualquier otra obra derivada. La comunicación y difusión al público por cualquier procedimiento o medio (impreso o electrónico). La inclusión en bases de datos y en sitios web, sean éstos onerosos o gratuitos, existiendo con ellos previo convenio perfeccionado con la Pontificia Universidad Javeriana Cali para efectos de satisfacer los fines previstos. En estos eventos, tales sitios tendrán las mismas facultades que las aquí concedidas para la referida universidad, con las mismas limitaciones y condiciones. El o los autores continúan conservando los correspondientes derechos sin modificación o restricción alguna, puesto que de acuerdo con la legislación colombiana aplicable, el acuerdo jurídico con la Pontificia Universidad Javeriana Cali, en ningún caso conlleva la enajenación del derecho de autor y de sus conexos. EL AUTOR, expresa que el artículo, folleto o libro objeto de la presente autorización es original y la elaboró sin quebrantar ni suplantar los derechos de autor de terceros, y de tal forma, el recurso electrónico aquí presentado es de su exclusiva autoría y tiene la titularidad sobre éste. PARÁGRAFO: en caso de queja o acción por parte de un tercero referente a los derechos de autor sobre el recurso electrónico en cuestión, EL AUTOR, asumirá la responsabilidad total, y saldrá en defensa de los derechos aquí autorizados; para todos los efectos, la Pontificia Universidad Javeriana Cali actúa como un tercero de buena fe.spa
dc.rights.urihttp://creativecommons.org/licenses/by-nc-nd/2.5/co/*
dc.subjectFacultad de Ingenieríaspa
dc.subjectIngeniería de Sistemas y Computaciónspa
dc.subjectOptimizationspa
dc.subjectEvolutionary Computationspa
dc.subjectNeuroevolutionspa
dc.subjectGenetic Algorithmspa
dc.subjectPeer-to-Peerspa
dc.subjectBlockchainspa
dc.subjectProof of Workspa
dc.subjectReinforcement Learningspa
dc.subjectDecentralized Networksspa
dc.subjectDistributed Computingspa
dc.subjectNeuroevolutionspa
dc.titleComputación Evolutiva Descentralizada de Modelo Híbrido usando Blockchain y Prueba de Trabajo de Optimizaciónspa
dc.typeinfo:eu-repo/semantics/masterThesisspa
dc.audiencePontificia Universidad Javeriana communityspa
dc.audienceResearchsspa
dc.audienceJournalistsspa
dc.audienceOtherspa
dc.contributor.roleConsultor de tesisspa
dc.coverageCali; Lat: 03 24 00 N degrees minutes; Lat: 3.4000 decimal degrees; Long: 076 30 00 W degrees minutes; Long: -76.5000 decimal degreesspa
dc.creator.degreeMagister en Ingenieríaspa
dc.creator.emailharveybc@ingeni-us.comspa
dc.publisher.facultyIngenieríaspa
dc.publisher.programMaestría en ingeniería con énfasis en Ingeniería de Sistemas y Computaciónspa
dc.pubplace.cityCalispa
dc.pubplace.stateValle del Caucaspa
dc.rights.accesoAcceso abiertospa
dc.rights.accessRightsinfo:eu-repo/semantics/openAccessspa
dc.rights.ccAtribución-NoComercial-SinDerivadas 2.5 Colombia*
dc.source.bibliographicCitationK. Stanley, "Evolving Neural Networks through Augmenting Topologies," Evolutionary Computation, vol. 10 , no. 2, pp. 99-127 , 2002.spa
dc.source.bibliographicCitationL. Wu, "Magnetic Resonance Brain Image Classification by an Improved Artificial Bee Colony Algorithm," Progress in Electromagnetics Research, vol. 116, pp. 65-79 , 2011.spa
dc.source.bibliographicCitationJ. Kennedy and R. Eberhart, "Particle Swarm Optimization," in Proceedings of IEEE International Conference on Neural Networks, 1995.spa
dc.source.bibliographicCitationM. Wa, "Genetic Algorithm and its application to Big Data Analysis," International Journal of Scientific & Engineering Research,, vol. 5, no. 1, 2014.spa
dc.source.bibliographicCitationP. Verbancsics, "Classifying Maritime Vessels from Satellite Imagery with HyperNEAT," in Proceedings of the Companion Publication of the 2015 on Genetic and Evolutionary Computation Conference - GECCO Companion, New York, USA, 2015.spa
dc.source.bibliographicCitationY.-J. Gong, "Distributed evolutionary algorithms and their models: A survey of the state-of-the-art.," Applied Soft Computing, 2015.spa
dc.source.bibliographicCitationJ. Lienig, "A parallel genetic algorithm for performance-driven VLSI routing," IEEE Trans. Evol. Comput, vol. 1, no. 1, pp. 29-39, 1997.spa
dc.source.bibliographicCitationH. Piereval, "Distributed evolutionary algorithms for simulation optimization," IEEE Trans Sys. Man. Cybern. , vol. 30, no. 1, pp. 15-24, 2000.spa
dc.source.bibliographicCitationK.C.Tan, "Automating the drug-sheduling of cancer chemotherapy via evolutionary computation," Artif.Intellig.Med., vol. 25, no. 2, pp. 169-185, 2002.spa
dc.source.bibliographicCitationJ. Creput, "Automatic mesh generation for mobile network dimensioning using evolutionary approach," Evol.Comput, vol. 9, no. 1, pp. 18-30, 2005.spa
dc.source.bibliographicCitationJ.Liu, "An evolutionary autonomous agents approach to image feature extraction.," IEEE Trans. Evol. Comput., vol. 1, no. 2, pp. 141-158, 1997.spa
dc.source.bibliographicCitationM. Epitropakis, "Hardware-friendly higher-order neural network training using distributed evolutionary algorithms.," Appl. Soft. comput. , vol. 10, no. 2, pp. 398-408, 2010.spa
dc.source.bibliographicCitationL. Kattan, "Distributed evolutionary estimation of of dynamic traffic origin/destination," in 13th IEEE Conference on Intelligent Transport Systems, 2010.spa
dc.source.bibliographicCitationJ. Noyima, "Ensemble classifier design by parallel implementation of genetic fuzzy rule selection for large datasets.," in IEEE Congress on Evolutionary Computation, 2010.spa
dc.source.bibliographicCitationF. Rainville, "DEAP: A Python Framework for Evolutionary Algorithms," in GECCO, 2012.spa
dc.source.bibliographicCitationM. Linder, "Grid computing in Matlab for solving evolutionary algorithms," in Technical Computing Bratislava, Bratislava, 2012.spa
dc.source.bibliographicCitationJ. Kennedy and R. Eberhart, "Particle Swarm Optimization," in Proceedings of IEEE International Conference on Neural Networks. , 1995.spa
dc.source.bibliographicCitationD. D. Karaboga, "An Idea Based On Honey Bee Swarm for Numerical Optimization," Erciyes University, 2005.spa
dc.source.bibliographicCitationM. Dorigo, "Distributed Optimization by Ant Colonies," in Première conférence européenne sur la vie artificielle, Paris, France, 1991.spa
dc.source.bibliographicCitationC. Huang, "A hybrid stock selection model using genetic algorithms and support vector regression," Applied Soft Computing, vol. 12, p. 807–81, 2012.spa
dc.source.bibliographicCitationE. Levy, "Genetic algorithms and deep learning for automatic painter classification," in conference on Genetic and evolutionary computation - GECCO ’14 , New York, New York, USA, 2014.spa
dc.source.bibliographicCitationE. Levy, "Genetic algorithms and deep learning for automatic painter classification," in Proceedings of the 2014 conference on Genetic and evolutionary computation - GECCO ’14 , New York, New York, USA, 2014.spa
dc.source.bibliographicCitationS. Whiteson, "Evolutionary Function Approximation for Reinforcement Learning," Journal of Machine Learning Research, vol. 8, pp. 877-917, 2006.spa
dc.source.bibliographicCitationR. B. Greve, "Evolving Neural Turing Machines for Reward-based Learning," in Genetic and Evolutionary Computation Conference (GECCO), Denver, Colorado, USA, 2016.spa
dc.source.bibliographicCitationA. Santoro, "One-shot Learning with Memory-Augmented Neural Networks," Cornell University , Ithaca, NY, USA, 2016.spa
dc.source.bibliographicCitationD. Izzo, "The generalized isalnd model," Studies in Computational Intelligence, vol. 415, pp. 151-169, 2012.spa
dc.source.bibliographicCitationG. Folino, "P-CAGE: An Environment for Evolutionary Computation in Peer-to-Peer Systems," in European Conference on Genetic Programming, Berlin, 2006.spa
dc.source.bibliographicCitationA. L. Ian Scriven, "Decentralised distributed multiple objective particle swarm optimisation using peer to peer networks," in IEEE World Congress on Evolutionary Computation, 2008. CEC 2008, 2008.spa
dc.source.bibliographicCitationA. L. C. Fernando Silva, "odNEAT: An Algorithm for Decentralised Online Evolution of Robotic Controllers," Evolutionary Computation , vol. 23, no. 3, pp. 421-449, 2015.spa
dc.source.bibliographicCitationD. Jakobović, "ECF - Evolutionary Computation Framework," University of Zagreb, http://ecf.zemris.fer.hr/, 2014.spa
dc.source.bibliographicCitationM. A. García-Sánchez P., "A Methodology to Develop Service Oriented Evolutionary Algorithms," Intelligent Distributed Computing VIII, vol. 570, 2015.spa
dc.source.bibliographicCitationS. Cahon, "ParadisEO: A Framework for the Reusable Design of Parallel and Distributed Metaheuristics," Journal of Heuristics, vol. 10, no. 3, pp. 357-380, 2004.spa
dc.source.bibliographicCitationD. R. White, "Software review: the ECJ toolkit," Genetic Programming and Evolvable Machines, vol. 13, no. 1, pp. 65-67, 2011.spa
dc.source.bibliographicCitationM. G. Arenas, "A Framework for Distributed Evolutionary Algorithms," Lecture Notes on Computer Science, vol. 2439, pp. 665-675, 2002.spa
dc.source.bibliographicCitationJ. L. J. Laredo, "Resilience to churn of a peer-to-peer evolutionary algorithm," International Journal of High Performance Systems Architecture, vol. 1, no. 4, pp. 260-268, 2008.spa
dc.source.bibliographicCitationC. Rohrs, "Query Routing for the Gnutella Network," Lime Wire LLC, 2002.spa
dc.source.bibliographicCitationR. Fielding, "Chapter 5: Representational State Transfer (REST)," in Architectural Styles and the Design of Network-based Software Architectures, Irvine, Ca, University of California, 2000.spa
dc.source.bibliographicCitationP. C. R. K. Len Bass, Software Architecture in Practice - Second Edition, Addison Wesley, 2003.spa
dc.source.bibliographicCitationS. Nakamoto, "Bitcoin: A Peer-to-Peer Electronic Cash Syste," bitcoin.org, vol. Retrieved 28 April 2016.spa
dc.source.bibliographicCitationH. C. M. Kalodner, "An empirical study of Namecoin and lessons for decentralized namespace design," in Workshop on the oconomics of information security, Delft, The Netherlands, 2015.spa
dc.source.bibliographicCitationD. Carboni, "Feedback based Reputation on top of the Bitcoin Blockchain," Retrieved from http://arxiv.org/abs/1502.01504 .spa
dc.source.bibliographicCitationE. Foundation, "Ethereum's white paper," Ethereum Foundation, 2014.spa
dc.source.bibliographicCitationG. N. O. Zyskind, "Decentralizing Privacy: Using Blockchain to Protect Personal Data," IEEE Security and Privacy Workshops , p. 180–184, 2015.spa
dc.source.bibliographicCitationS. Nakamoto, "Bitcoin: A Peer-to-Peer Electronic Cash System," in Web. bitcoin.org. Retrieved 28 April 2014, 2008.spa
dc.source.bibliographicCitationS. King, "Primecoin: Cryptocurrency with Prime Number Proof-of-Work," 2013 .spa
dc.source.bibliographicCitationA. R. Marshall Ball, "Proofs of Useful Work," Cryptology ePrint Archive, 2017.spa
dc.source.bibliographicCitationR. A. G. Barto, Reinforcement Learning: An Introduction, Cambridge, Massachusetts: The MIT Press, 2005.spa
dc.source.bibliographicCitationM. S. J. Moody, "Learning to trade via direct reinforcement," IEEE Transactions on Neural Networks, vol. 12, no. 4,, 2001spa
dc.source.bibliographicCitationC. WATKINS, "Q,-Learning," Machine Learning, vol. 8, 1992.spa
dc.source.bibliographicCitationR. M. Kenneth O. Stanley, "Efficient Reinforcement Learning through Evolving Neural Network Topologies," in Genetic and Evolutionary Computation Conference , San Francisco, CA, 2002.spa
dc.source.bibliographicCitationC. Igel, "Neuroevolution for reinforcement learning using evolution strategies," in Congress on Evolutionary Computation, 2003.spa
dc.source.bibliographicCitationV. Heidrich-Meisner, "Neuroevolution strategies for episodic reinforcement learning," Journal of Algorithms, vol. 64, no. 4, pp. 152-168, 2009.spa
dc.source.bibliographicCitationD. Lessin, "Open-ended behavioral complexity for evolved virtual creatures," in 15th annual conference on Genetic and evolutionary computation, Amsterdam, The Netherlands, 2013.spa
dc.source.bibliographicCitationG. B. a. V. Cheung, "OpenAI Gym," Arxiv, vol. arXiv:1606.01540, 2016.spa
dc.source.bibliographicCitationN. G. L. Iker Zamora, "Extending the OpenAI Gym for robotics: a toolkit for reinforcement learning using ROS and Gazebo," Whitepaper, 2016.spa
dc.source.bibliographicCitationA. H. David Silver, "Mastering the game of Go with deep neural networks and tree 62 search," Nature, vol. 529, no. 7587, pp. 484-489, 2016.spa
dc.source.bibliographicCitationK. K. Volodymyr Mnih, "Human-level control through deep reinforcement learning," Nature, vol. 518, no. 7540, pp. 529-533, 2015.spa
dc.source.bibliographicCitationU. B. o. E. Analysis, "U.S. INTERNATIONAL TRADE IN GOODS AND SERVICES - December 2013," U.S. Department of Commerce, Washington, DC 20230, 2013.spa
dc.source.bibliographicCitationE. Department, "Triennial Central Bank Survey of foreign exchange and OTC derivatives markets in 2016," Bank For International Settlements, Basel, Switzerland, http://www.bis.org/publ/rpfx16fx.pdf - Retrieved on 20 March of 2017.spa
dc.source.bibliographicCitationInvestopedia, "Forex Broker Definition," http://www.investopedia.com/terms/forex/c/currency-trading-forex-brokers.asp, Retrieved on March 20 of 2017.spa
dc.source.bibliographicCitationInvestopedia, "Margin Call Definition," http://www.investopedia.com/terms/m/margincall.asp, Retrieved on March 3 of 2017.spa
dc.source.bibliographicCitationInvestopedia, "What is a Spread," http://www.investopedia.com/terms/s/spread.asp, Retrieved on March 3 of 2017.spa
dc.source.bibliographicCitationInvestopedia, "Leverage Definition," http://www.investopedia.com/terms/s/spread.asp, Retrieved on March 20 of 2017.spa
dc.source.bibliographicCitationA. P. CHABOUD, "Rise of the Machines: Algorithmic Trading in the Foreign Exchange Market," Journal of the American Finance Association, vol. 69, no. 5, p. 2045–2084 , 201.spa
dc.source.bibliographicCitationM. A. H. Dempster, "An automated FX trading system using adaptive reinforcement learning," Expert Systems with Applications, vol. 30, no. 3, pp. 543-552, 2006.spa
dc.source.bibliographicCitationHistData.com, "Free Forex Historical Data Repository," http://www.histdata.com/, Retrieved on Mach 20 of 2017.spa
dc.source.bibliographicCitationR. T. Fielding, "Chapter 5: Representational State Transfer (REST)," in Architectural Styles and the Design of Network-based Software Architectures, Irvine, University of California, 2000.spa
dc.source.bibliographicCitationH. W. Group, " SOAP: Simple Object Access Protocol," [Online]. Available: https://tools.ietf.org/html/draft-box-http-soap-00.spa
dc.source.bibliographicCitatione. a. Don Box, "Simple Object Access Protocol (SOAP) 1.1," W3C, 8 May 2000. [Online]. Available: https://www.w3.org/TR/2000/NOTE-SOAP-20000508/. [Accessed 18 04 2018].spa
dc.source.bibliographicCitationJ.-R. W. Group, "JSON-RPC 2.0 Specification," JSON-RPC Working Group, 04 01 2013. [Online]. Available: http://www.jsonrpc.org/specification. [Accessed 18 04 2018].spa
dc.source.bibliographicCitationR. Mikkulainen and J. e. a. Liang, "Evolving Deep Neural Netwokrs," eprint arXiv:1703.00548, 2017.spa
dc.source.repositoryReponame:Vitela: Repositorio Institucional PUJspa
dc.source.institutionInstname:Pontificia Universidad Javeriana Calispa
dc.type.hasversioninfo:eu-repo/semantics/publishedVersionspa
dc.type.spaTesis Maestríaspa


Files in this item

Thumbnail
Thumbnail
Thumbnail

This item appears in the following Collection(s)

Show simple item record

Atribución-NoComercial-SinDerivadas 2.5 Colombia
Except where otherwise noted, this item's license is described as Atribución-NoComercial-SinDerivadas 2.5 Colombia