Research on the Pricing Method of Automotive Aftermarket Parts Based on Distributed Computing Architecture

Abstract

With the rapid development of the automobile industry, the scale of the automobile aftersales parts market continues to expand, and its pricing is of great significance to both enterprises and consumers. The traditional pricing method has many limitations, it is difficult to adapt to the dynamic changes of the market. In this paper, an automatic pricing tool for automotive aftermarket parts based on distributed computing architecture is deeply studied, aiming to improve the accuracy and efficiency of pricing. Firstly, the cost and sales price functions of auto aftersales parts are determined. Based on the characteristics of discrete and multi-dimensional data of auto aftersales parts, new features are constructed by using feature engineering method, including data acquisition and preprocessing, model architecture design, etc. Then, it uses the calculation principle of Spark and Spark Shuffle to build distributed computing system and distributed training strategy. Finally, the model is evaluated and optimized, and the existing risks and countermeasures are proposed. The method in this paper can effectively solve many problems faced by the traditional pricing, and improve the accuracy and real-time pricing. It will bring significant economic and social benefits to the enterprise, and promote the healthy development of the automobile aftermarket.