Font Size: a A A

A Study On Pricing Model Of Fresh Commodities In Supermarket

Posted on:2020-07-25Degree:MasterType:Thesis
Country:ChinaCandidate:F Y FanFull Text:PDF
GTID:2428330623960343Subject:Applied statistics
Abstract/Summary:PDF Full Text Request
Nowadays,with the continuous improvement of people's living standards,people's daily needs are quite different from those in the past.They begin to pursue the quality of life.For fresh commodities,more and more people choose to go to supermarkets or special fresh franchise stores to buy.Therefore,in recent years,the management status of fresh commodities in supermarkets has been constantly improved,and some supermarkets and stores which have done better in fresh commodities have been born.Therefore,in the pricing of fresh commodities,how to set a reasonable price to attract more customers and get more profits is a problem that most operators need to consider.If we can use some research methods and historical data to transform supermarket staff's experience in fresh commodities pricing into a special model,that is,input some known information to export the final price of commodities,we can save the trouble of manual pricing and improve work efficiency.The purpose of this paper is to find out the characteristics that affect the price change by analyzing the historical transaction data of fresh commodities in a supermarket in Fuzhou in the first half of 2017,and to establish the pricing model of fresh commodities by using various research methods.The main work includes the following aspects:1)We use these trading data to analyze the regularity of commodity price changes and the factors that may affect price changes.At the same time,we carry out feature analysis including the test of outlier,missing value filling.2)After choosing the appropriate features,the methods of general linear regression,logical regression,K-nearest neighbor,random forest and XGBOOST are used by us to model and forecast the price of fresh commodities,and we use some testing methods to evaluate the performance of the model.The results show that the linear regression method is not suitable for predicting the price of fresh commodities.Relatively speaking,other machine learning methods can achieve better prediction results.3)Because the price of fresh commodities is a time series(arranged in sequence according to the time of sale of commodities),this paper also uses ARIMA model and Multiplicative Seasonal model to analyze and model,and the analysis results show that the fitting effect of the model considering periodic effect is better.
Keywords/Search Tags:Fresh commodities price, Feature analysis, Machine learning, Time series
PDF Full Text Request
Related items