The Analysis Of Hotel Price Prediction Based On Machine Learning

With the rapid development of information technology and high-speed mobile internet,more and more people tend to choose travel products online.The travel industry grows quickly.It is reasonable to improve consumers' experience by providing a hotel price prediction between the search time and the check-in time.Due to the dynamics of hotel prices,it is challenging to predict the price with high accuracy.This paper provides a novel method to predict the hotel price based on the inherent characters and historic price.The contributions of the work are summarized as follows:1)We crawl the hotel price information for the 18 of most popular cities,including the prices between current day and the day 8th days later,and the basic information which contains the opening year,location information etc.We analyze the hotel characteristics and the price flutruations between the different types of hotels.2)Applying the classic machine learning methods,we predict the hotel price and get the factors which may affect the price much.3)Based on the ensemble learning algorithm,we further propose a Hierachrical Ensemble Learning Model(HELM)to predict hotel price.Experimental results validate the efficiency of the proposed methods.