Font Size: a A A

Research And Development Of Data Analysis System For Anti-counterfeiting Traceability

Posted on:2018-07-29Degree:MasterType:Thesis
Country:ChinaCandidate:G M ZhaiFull Text:PDF
GTID:2428330596954214Subject:Computer technology
Abstract/Summary:PDF Full Text Request
The national product anti-counterfeit traceability verification platform accumulates a large number of historical query data.It is of great significance to design and implement the data analysis system by applying data mining technology to discover the pattern of anti-counterfeiting historical query data.This thesis includes the following works:(1)Respectively,the geographical location,time,product categories and user inquiry means are used as the primary query conditions for statistical analysis on data collection.Then ECharts is used to display analysis results.(2)In order to improve the prediction accuracy of time series data with BP neural network,the wavelet analysis and BP neural network model are combined.Firstly,the original data is denoised by wavelet.Then,the denoised data is decomposed in multiple layers with wavelet analysis.The decomposed data is used for learning intelligent model,combined with forecasting method to predict the abnormal query data and the public safety event.(3)Since the K-means clustering method is sensitive to initial clustering centers and the poor global search ability,a K-means clustering algorithm based on Artificial Bee Colony is proposed.Firstly,in order to solve the problem of the slow convergence of Artificial Bee Colony algorithm,gravitational mass calculation is employed to replace the roulette probability selection and to reduce the time complexity.Then,clustering analysis of product®ion objects is carried out by improved clustering method.Finally,the value model is constructed by using analytic hierarchy process.Based on the value model,the value of the cluster is estimated,which provides a reference for the marketing of the enterprise.With these methods above,the data analysis system based on hierarchical software architecture is implemented by using the.NET development framework and Oracle database management system.This data analysis system is integrated into the national product anti-counterfeit traceability verification platform.
Keywords/Search Tags:Statistical Analysis, BP Neural Network, Time Series Prediction, K-means Clustering Algorithm, Object Value Clustering
PDF Full Text Request
Related items