联系我们

论文查询结果	返回搜索

论文编号：	12761
作者编号：	2320190339
上传时间：	2021/12/8 15:56:02
中文题目：	汽车品牌客户保留预测模型构建及应用研究
英文题目：	Research on Construction and Application of Automobile Brand Customer Retention Forecast Model
指导老师：	李东进
中文关键字：	汽车品牌；客户保留；主成分分析；机器学习；预测模型
英文关键字：	Automobile brand; Customer retention; Principal component analysis; Machine learning; Forecast model
中文摘要：	2001年加入世界贸易组织（WTO）后，中国汽车产业经历了高速发展阶段，自2009年以来，中国已连续11年成为世界第一大汽车市场。但是，自“十三五”以来中国汽车市场增长速度已呈现逐渐放缓趋势，基于中国经济发展态势、人口增长变化以及城市交通状况等综合因素，未来高速增长特征将难以重现。同时，随着汽车市场保有量的不断增加以及换购周期的到来，换购将成为带动新车市场销售的新动力。从增量市场竞争向存量市场竞争的变化成为中国未来汽车产业竞争发展的新趋势。在此背景下，有效掌握保有客户在换购中影响其决策的相关信息、提升换购中客户保留比例，成为各汽车企业关注的核心问题之一。汽车品牌客户保留的合理预测对于车企在制定和优化其营销策略过程中具有十分重大的意义。 本论文以汽车企业对客户换车过程中保留情况预测的实际需求为出发点，重点解决传统统计分析方法与经验判断无法有效处理复杂多样的消费者换车决策过程信息的问题，创新性的将机器学习与客户保留预测相结合，以中国合资品牌客户换车过程调研数据为基础，分别尝试用bagging、boosting和deep learning三种算法完成汽车品牌客户保留预测模型的训练与测试。基于模型预测结果完成模型对比分析，选择性能表现最优模型作为最终汽车品牌客户保留预测模型，最后依靠此模型完成具体汽车品牌调研客户换购过程的保留情况预测与提升策略建议。 通过本论文得出以下结论：首先，存量竞争趋势下汽车企业客户保留情况的预测对其制定营销策略具有重要意义；其次，人工智能和机器学习等新兴方法与传统市场营销的结合可以有效用于消费者购买决策过程中复杂信息的处理；再次，以机器学习方法结合汽车换购客户相关调研数据可以实现客户保留预测模型的训练；最后，本论文所构建预测模型可以有效应用于汽车品牌换购客户保留情况的预测，并基于预测结果为客户保留提升策略提出建议。
英文摘要：	After the entrance of WTO in 2001, China’s automobile industry has experienced a stage of rapid development. Since 2009, China has become the largest automobile market in the world for 11 years. However, after the “13th Five-Year Plan”, the growth of China’s automobile market has slowed down and the fast growth momentum will not reappear on the basis of the comprehensive factors such as China’s economic development trend, population growth change, and urban traffic conditions. Meanwhile, with the continuous increase of the market ownership and the advent of the replacement cycle, the replacement of new cars will become a new driving force. The change from incremental market competition to stock market competition has become a new trend of China’s future auto industry competition. Under this circumstance, it has become one of the core issues concerned by automobile companies is to get useful information that affects the retained customer’s decision-making and increase the proportion of customer retention during the exchange purchases. The reasonable forecasts of automobile brand customer retention are of great significance for automobile companies in the process of formulating and optimizing their marketing strategies. This thesis starts from the actual demand of customer retention forecast that automobile companies made during the process of customers changing their cars, and then focuses on solving the problem that both traditional statistical analysis methods and empirical judgment cannot effectively deal with the complex and diverse information of the consumers’ decision-making process when changing their cars, and finally combine the machine learning with customer retention forecast innovatively and on basis of the investigation data of Chinese joint-venture brand customers’ process of changing cars, try to use bagging, boosting and deep learning algorithms to complete the training and testing of automobile brand customer retention forecast model. According to the forecast results and the model comparative analysis, choose the best performance model as the final car brand customer retention forecast model, and then use the model to complete the retention forecast of the specific car brand research on customer exchange process and put forward the improved strategies. The conclusions of this thesis are as follows: firstly, under the trend of stock competition, the forecast of customer retention is of great significance for automobile companies to formulate marketing strategies; secondly, the combination of new methods such as artificial intelligence and machine learning with traditional marketing methods can effectively deal with the complex information in the process of consumers’ purchase decision-making; thirdly, the training of customer retention forecast model can be realized by using machine learning method combined with the relevant survey date of automobile replacement customers; finally, the forecast model in this thesis can be effectively applied to the specific car brands customer retention forecasts, and give suggestions for customer retention improvement strategies on the basis of the forecast results.
查看全文：	预览  下载(下载需要进行登录)