The Photo Task Pricing Based on Multi-objective Programming Model

Abstract

This thesis establishes the functional relationship between the position, pricing and completion of the task, the position, reputation value of the member, the starting time and the reservation limit of the task given by referring to its reputation. In addition, this thesis also establishes multi-objective programming model and logistic regression model, and writes out the task pricing rule and designs the pricing scheme, realizes the model of obtaining the best pricing scheme by analyzing the task information and member information. Firstly, the data in the attachment are plotted, the outliers are removed and the error values are modified. In addition, by plotting the pre-processed data, we can extract the characteristics of the region and find out the various factors that affect the pricing law and the general relationship. At the same time, on the basis of theoretical analysis and data availability, this paper establishes a task pricing model by using multiple binomial regression algorithm, obtains the law of task pricing in Dongguan, Shenzhen and Guangzhou, and analyzes the reasons for the low completion rate of tasks. In addition, considering from the point of view of task publisher, we should not only make the order rate sufficiently high, but also make the order cost not too high. On this basis, the thesis puts forward a minimum-maximum bi-objective optimization model. In order to solve this model, this paper creatively puts forward the cost performance function, and designs the corresponding solving algorithm to solve the problem and give the pricing scheme. Finally, based on the consideration of jointly packaging and publishing tasks with similar distances, the multi-objective optimization model is adjusted, and two restrictions are given: the distance between packaged tasks is less than a fixed value, and the number of packaged tasks is less than a fixed value.