Font Size: a A A

Mining Forecasting And Visualization Of Pig Market Price Data

Posted on:2019-04-10Degree:MasterType:Thesis
Country:ChinaCandidate:X X DongFull Text:PDF
GTID:2428330548979788Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
With the continuous improvement of the informatization level of pig breeding industry in our country and the continuous development of Internet industry,a great deal of price data has been accumulated in the transactions of pig market.The price data is of great significance to guide the development of pig industry.However,these data come from a number of different sources which have huge differences in data quality,together with the huge differences in data richness between different regions poses great challenges in using these price data.Based on the above challenges,this paper proposes the following solutions:First,aiming at the problem of how to use the data with huge differences in quality to generate the representative prices of different regions,this paper proposes a price collection model based on artificial neural network.Comprehensive using the data from various sources,automatically learn the proper source weight,generate a representative regional prices.On the other hand,aiming at the region where data richness is extremely lacking,this paper proposes a vacancy price filling model based on SVD.The model rationalizes the vacancy price by studying the price relation among different regions and the history of price changes in the same region,solve the problem of the extreme lack of prices in some areas.In the meantime,this paper uses RNN(Recurrent Neural Network)to forecast the future pig price,which further enhances the guiding significance of the pig market data to farmers and sellers.Finally,this paper designs and implements the live pig price visualization system,which provides users with a more convenient and intuitive way of price acquisition based on various visual forms such as maps and line charts.This visualization system uses message oriented middleware,distributed file system,distributed coordination and other system components to decouple the system,to further ensure the system performance,availability and scalability.Based on the pig industry informatization development status of our country,this paper presents a series of models for generating pig prices,which provides the main reference for farmers and other market participants to arrange procurement,production and sales planning reasonably and reduce blindness.
Keywords/Search Tags:Pig price, Neural network, Incidence relation, Price filling, Price forecasting, Data visualization
PDF Full Text Request
Related items