Design And Implementation Of Big Data Analysis And Forecast System Based On Deep Learning

With the rapid development of information and communication technol-ogy,the amount of global data has exploded.How to effectively analyze the complicated data and use its value reasonably is one of the important topics at present.In the fields of speech recognition,image understanding,natural lan-guage processing,video recommendation and other data processing and appli-cation,deep learning not only changes the traditional machine learning methods and has also triggered a breakthrough change.The flourishing development of deep learning has made the need for deep learning applied to various industries more urgent.However,deep learning has a relatively high threshold for users in the general industry.It requires a lot of time cost for learning and complex underlying model coding,and also needs expensive hardwares to build deep learning frameworks.Based on the above research background,this paper designs and imple-ments a big data analysis and forecast system based on deep learning,which aims to support deep learning,reduce user cost and operational complexity.Firstly,this paper investigates the existing processing frameworks and related technologies related to big data and deep learning.Combined with the analysis of design scenarios,functional requirements and performance requirements,it designs the five-layer architecture,including data absorption,platform frame-work,algorithm model,integration components and services application.It lays the foundation for platform specific implementation.Then the design and implementation goals of each functional module of the system are introduced in detail.In addition,in order to improve the training efficiency of the algorithm model used in the platform,this paper studies and analyzes the convolutional neural network.It proposes a new distributed convolutional neural network parallelism strategy,which optimizes the convolutional neural network model based on the distributed parallel processing framework.Finally,it proves the design and implementation of the platform meets the functional and performance requirements of the platform design goals by test analysis.