With the emergence of new retail, instant distribution has a wider space of development, consumers also put forward higher requirements for it. In order to study the instant distribution demand under new retail, based on the consumption the takeaways demand in Xi’an, this study forecasts the demand with qualitative and quantitative methods. The data mining of takeaways consumption demand in Xi’an is analyzed by using online questionnaire survey. This paper analyzes consumer behavior and the spatial and temporal distribution of instant distribution demand, provides guiding suggestions for the takeaway merchants by analyzing the survey results. Based on the practicalities of the take-out market in China, the forecasting model of instant distribution demand which represented by the takeaways consumption demand in Xi’an, which is trained and simulated with MATLAB software, is established by using Multiple Regression Analysis and Back Propagation Neural Network. The results shows that the takeaway products value varies greatly among different groups of people, the sales volume of businesses with high scores in all aspects remains good, while that of businesses with low scores in some aspects is relatively volatile. The results has reference value to guide the takeaway merchant how to improve the sales of their product. KEY WORDS: new retail, instant distribution, Multiple Regression Analysis, BP Neural Network, prediction

CICTP