User travel intention and type prediction method based on gradient improvement decision tree

Abstract

A user travel intention and type prediction method based on a gradient improvement decision tree comprises the following steps: step 1, collecting user basic information, service use, family network intimacy and a travel behavior data table provided by a mobile operator, and performing desensitization; step 2, performing data preprocessing on the data table; step 3, constructing vectors to represent corresponding categories according to the number of the travel categories in the sample data; step 4, training a classification regression tree for each possible category in the sample data; step 5, calculating a loss function value of each feature on each feature value, and constructing a prediction function under the condition that the loss function value is minimum; and step 6, generating detailed data of the to-be-predicted user on each feature, and performing prediction by using a prediction function. According to the method, the travel intention of a user and the type of the destination are predicted by using the gradient improvement decision tree algorithm, and the accuracy is relatively high.

Description

technical field [0001] The invention relates to a user travel intention and type prediction method based on a gradient-lifting decision tree. [0002] technical background [0003] With the development of economy and technology, the improvement of transportation facilities and the popularization of the Internet have made people's travel more convenient. More and more people go out to travel, and there are many types of destinations for people to travel. For short-distance travel, some people like to travel outside the province, while others choose to travel abroad or to foreign attractions. [0004] Before traveling, most people will ask relatives and friends or search on the Internet to understand the situation of the tourist destination, make a strategy, and arrange the travel itinerary reasonably. Therefore, it is possible to predict whether people have travel intentions and the type of destination through their daily calls or online behaviors, and help tourism practition...

AI Technical Summary

Problems solved by technology

[0006] In order to overcome the shortcomings of the poor accuracy of the existing artificial phone calls or questionnaire survey methods, the present invention uses data such as user basic information, business use, family network intimacy and travel behavior provided by mobile operators, and after data preprocessing Use the gradient boosting decision tree algorithm to predict the user's travel intention and destination type with high accuracy

Images