Genetic Algorithm Research And Application Based On Hadoop

With the rapid development of Internet,our life has been being improved increasingly. Atthe same time, produced information data has been increasing by the speed of the exponential.In the scene, Big Data occurs. We can use its theory and techniques to process massive data,in order that better services are given to people and the development and improvement ofmarket economy is promoted rapidly. The solving of many problems of the most fields ofsicence has shifted to the proceesing of Big Data, such as meteorological data mining andweather forecast, the processing of Internet text files and searching index, astronomicalresearch, genetic recombination,military reconnaissance, the large-scale facility location. Themost of those can be solved by genetic algorithm.But more massive data must be processed tosolve those than before. Tranditional genetic algorithm can not solve those. The occurrence ofCloud Computing results in the new life of Big Data, especially Hadoop is used to be verygood solution of Big Data processing.As to the problems that can been solved by genetic algorithm in Big Data, geneticalgorithm research has profound significance. To meet the need of social development, mangresearchers have been doing a lot of work in thought and theory of that and come up withspecific thought and theory. But the thought and theory has been in front stage. And a lot ofpractice is done to improve the thought and theory.In consequence, continue to researching how genetic algorithm solves problems in BigData and make it to move a forward single step. There is application value and practicalsignificance. In the thesis, firstly read a lot of papers about the implementation rationale,model and specific implementation of genetic algorithm in Hadoop. Summarise twoimplementation models of genetic algorithm in Hadoop, vertical strategy and horizontalstrategy.And deeply study and analyse the two implementation models, their design andimplementation. Summarise their merit and demerit. Based on above, come up with improvedimlpementation model of genetic algorithm in Hadoop, its detailed design and generalimplementation steps. Data structure is designed, which is used in the communication witheach stage of improved implementation model in the thesis.Hadoop cluster is builded as a testplatform. The improved implementation model in this thesis is tested by test data to verify itscorrectness and validity. Exprimental results prove to be that the improved implementationmodel of this thesis is correct and valid.Finally, the improved implementation model of thisthesis is applied to solving facility location of emergency system.