基于线性回归和神经网络模型的二手车交易价格预测分析  

Prediction and analysis for second-hand car transaction price based on linear regression and neural network models

作　　者：郑爱萍 李彬彬[1] 郭传好[1,2] ZHENG Aiping;LI Binbin;GUO Chuanhao(School of Economics and Management,Zhejiang Sci-Tech University,Hangzhou 310018,China;Zhejiang Academy of Ecological Civilization,Zhejiang Sci-Tech University,Hangzhou 310018,China)

机构地区：[1]浙江理工大学经济管理学院,杭州310018 [2]浙江理工大学浙江省生态文明研究院,杭州310018

出　　处：《智能计算机与应用》2023年第9期103-110,共8页Intelligent Computer and Applications

基　　金：浙江理工大学研究生教育教学改革研究项目(YJG-M202008);2021年度浙江省教育厅一般科研项目(Y202148252);广州市科学技术局2020年科技创新发展专项资金(202002030119)。

摘　　要：二手车市场发展和需求潜力巨大,二手车价格的制定是影响和制约二手车市场发展的一个重要因素,因此构建一个合理有效的二手车交易价格预测模型十分亟需和重要。本文以58同城二手车交易平台数据为基础,分析相关数据的含义及其关联性并对数据进行清洗与处理,分别建立多元线性回归和神经网络二手车交易价格预测模型,并以平均相对误差,准确率等评价指标评估分析了模型的性能。研究结果表明:基于对数归一化处理数据的预测模型具有更高的精度和稳定性;与多元线性回归预测模型相比,神经网络预测模型具有更好的精度和稳定性。相关研究成果对于指导二手车交易价格的合理制定和二手车市场的健康良性发展具有重要的意义。The development and demand of second-hand car market has great potential,and the price of second-hand car is an important factor that affects and restricts the development of second-hand car market.Therefore,it is urgent and important to build a reasonable and effective used car transaction price prediction model.Based on the data of 58.com used car trading platform,the meaning and correlation of relevant data were analyzed,and the data were cleaned and processed.Multiple linear regression and neural network used car trading price prediction models were established respectively,and the performance of the models was evaluated and analyzed with the average relative error,accuracy and other evaluation indicators.The results show that the prediction model based on logarithmic normalization has higher accuracy and stability.Compared with the multiple linear regression prediction model,the neural network prediction model has better accuracy and stability.The relevant research results are of great significance to guide the reasonable formulation of second-hand car trading prices and the healthy development of second-hand car market.

关 键 词：二手车交易 价格预测 线性回归 神经网络 平均相对误差 

分 类 号：F426.471[经济管理—产业经济] F724.6[自动化与计算机技术—控制理论与控制工程] TP183[自动化与计算机技术—控制科学与工程]