数据与计算发展前沿 ›› 2023, Vol. 5 ›› Issue (6): 115-125.

CSTR: 32002.14.jfdc.CN10-1649/TP.2023.06.011

doi: 10.11871/jfdc.issn.2096-742X.2023.06.011

• • 上一篇    下一篇

基于TPA-Transformer的机票价格预测

申志豪1,2(),李娜1,尹世豪4,5,*(),杜一1,2,3,胡良霖1,2,3   

  1. 1.中国科学院计算机网络信息中心，北京 100083
    2.中国科学院大学，北京 100049
    3.国家基础学科公共科学数据中心，北京 100083
    4.中国民航信息网络股份有限公司，北京 101318
    5.民航旅客服务智能化应用技术重点实验室，北京 101318
  • 收稿日期:2022-05-20 出版日期:2023-12-20 发布日期:2023-12-25
  • 通讯作者: 尹世豪（E-mail: shyin@travelsky.com.cn）
  • 作者简介:申志豪，中国科学院计算机网络信息中心，硕士研究生，主要研究方向为机器学习、时间序列预测。
    本文承担工作为：TPA-Transformer模型架构和实验测试。
    SHEN Zhihao is a master's student at the Computer Network Information Center, Chinese Academy of Sciences. His main research interests are machine learning and time series forecasting.
    In this paper, he is mainly responsible for experiments and tests of the TPA-Transformer model.
    E-mail: shenzhihao@cnic.cn|尹世豪，中国民航信息网络股份有限公司，工程师，学士，主要研究方向为人工智能。
    本文主要承担工作为：数据清洗、数据处理、特征筛选与处理。
    YIN Shihao, the bachelor, is an engineer at Travelsky Technology Limited. His main research interest is artificial intelligence.
    In this paper, he is mainly responsible for data cleaning, data processing, feature screening, and processing.
    E-mail: shyin@travelsky.com.cn
  • 基金资助:
    国家自然科学基金重点项目(61836013);北京市科技新星计划(Z191100001119090)

Airfare Price Prediction Based on TPA-Transformer

SHEN Zhihao1,2(),LI Na1,YIN Shihao4,5,*(),DU Yi1,2,3,HU Lianglin1,2,3   

  1. 1. Computer Network Information Center, Chinese Academy of Sciences, Beijing 100083, China
    2. University of China Academy of Sciences, Beijing 100049, China
    3. National Basic Science Data Center, Beijing 100083, China
    4. Travelsky Technology Limited, Beijing 101318, China
    5. Key Laboratory of Intelligent Passenger Service of Civil Aviation, Beijing 101318, China
  • Received:2022-05-20 Online:2023-12-20 Published:2023-12-25

RichHTML

19

PDF

298

摘要：

【目的】 航空业被认为是使用最复杂定价策略的行业之一，机票价格动态波动显著。乘客根据经验购票（如尽早购票）往往得不到最优价格，而基于时序模型的传统价格预测方法也不能很好地捕捉复杂内外部因素与机票价格之间的依赖关系。【方法】 本文提出基于机器学习的机票价格预测模型TPA-Transformer（Ticket Price Aware Transformer，机票价格感知Transformer）和一种基于时间序列的数据处理方法，通过增加注意力模块引入其他航班价格参考信息，并在Encoder后增加多层卷积结构进行多航班不同属性信息融合与局部特征提取，从而提高模型在多步价格预测上的表现性能。【结果】 随后在5个回归评价指标（MSE、RMSE、MAE、ACC以及AMS）上对模型结果进行验证。【结论】 结果证明，模型能够有效提高预测准确率，明显优于其他5个对比模型（随机森林、XGBoost、LSTM、GRU、Transformer）。

关键词: 机票价格预测, 机器学习, 时间序列, 注意力机制

Abstract:

[Objective] The airline industry is one of the industries that use the most complex pricing strategies. Even the ticket prices on the same flight fluctuate dynamically and significantly. Passengers can only make decisions based on experience, such as buying tickets as soon as possible, but it is not always the best choice. The traditional price forecasting methods cannot sufficiently capture the dependence between complex internal/external factors and the ticket prices. [Methods] This paper designs and implements the TPA-Transformer (Ticket Price Aware Transformer) to predict ticket price and proposes a related data processing method based on time series. This model adds the attention module to introduce reference information and the multi-layer convolution after the encoder to fuse the information of different features of multiple flights and extract local features to improve the model's performance in multi-step price prediction. [Results] The model is verified on five regression evaluation indexes (MSE, RMSE, MAE, ACC, and AMS). [Conclusions] Experiments show that the model effectively improves the prediction accuracy and is superior to other five comparison models (Random Forest, XGBoost, LSTM, GRU, and Transformer).

Key words: airfare price prediction, machine learning, time series, attention mechanism