Application Research Of Passenger Car Sales Forecasting Model Based On Search Index And Machine Learning

In recent years,China’s automobile production and sales are decreasing year by year,and the impact of external environment on China’s automobile industry is becoming more and more serious.Since 2018,the number of automobile production and sales,and some important economic indicators of the automotive industry have also shown a negative growth trend.The first half of 2020,COVID-19 has seriously affected the market economy including the automotive industry.From the perspective of the types of automobile products,the negative growth mainly occurred in passenger cars.Therefore,the analysis of the current situation of domestic passenger car sales market and the prediction of passenger car sales by reasonable methods can provide certain guidance for the government to evaluate the impact of the epidemic on the passenger car market,guide the recovery of the passenger car industry after the epidemic,and formulate passenger car sales policies.For the passenger car manufacturers,their strategic decisions are based on the demand of the passenger car market,and the sales volume of passenger cars is the best indicator to reflect the market demand.Therefore,the forecast of passenger car sales can analyze the demand of the passenger car market,and guide automobile manufacturers to make production and sales plans,which has important guiding significance for the future development planning of enterprises.This paper mainly studies the sales volume of domestic passenger car market.This paper first analyzes the current situation and development trend of China’s passenger car market,and then crawls the search index data of relevant keywords based on the consumer behavior that consumers in China’s passenger car market will collect information through search engines before buying passenger cars.Based on the collected search index data,the keywords are selected according to whether there is time difference leading and time difference correlation coefficient.On this basis,a passenger car sales forecasting model based on neural network and gradient promotion regression method is established.The model is integrated according to the model group and method with the least sum of residual squares,and a combined forecasting model of passenger car sales is established.The results show that the combination forecasting model based on machine learning algorithm and search index can not only reflect the change of passenger car sales timely and accurately,but also has better forecasting accuracy than the traditional time series forecasting model,and achieves ideal forecasting effect.As the passenger car sales forecasting model based on search index and machine learning algorithm is established on the real-time search index data,the model can also be better applied when taking the sales of brand passenger cars and various models of passenger cars as the research objects.