| 標題 Title |
澳門旅遊需求研究——基於機器學習及互聯網搜索熱度 Research on Macao’s Tourism Demand Based on Machine Learning and Internet Search Index |
|---|---|
| 作者 Author |
王俊皓，苑皓，阮智豪 WANG Junhao, YUAN Hao, RUAN Zhihao |
| 摘要 Abstract |
本研究提出了四種模型，分別是傳統的時間序列模型 ARIMA 以及機器學習類 模型隨機森林、LSTM 以及 CNN-LSTM，分別對澳門 2008 年 1 月至 2019 年 12 月的入境總 人數序列進行分析，並基於互聯網搜索熱度數據（百度指數、谷歌趨勢），比較不同的搜 索引擎熱度代表的海內外遊客搜索偏好的數據。研究結果表明了在結合谷歌和百度搜索熱 度下的 CNN-LSTM 模型的性能最優，且 CNN-LSTM 模型並不對數據的預處理過程有嚴格 的要求，更加適合本研究的場景，即澳門旅遊需求預測。
This study proposes four models, namely, the traditional time series model ARIMA, the machine learning model random forest, LSTM and CNN-LSTM. We use them to analyze the sequence of the total number of immigrants in Macao from January 2008 to December 2019. Based on Internet search popularity data (Baidu Index, Google Trends), compared with the data of domestic and overseas tourists’ search preferences represented by the popularity of different search engines. The research results show that the performance of the CNN-LSTM model combined with the search popularity of Google and Baidu is the best, and the CNN-LSTM model does not have strict requirements on the data preprocessing process, which is more suitable for the scene of this study, namely the tourism demand of Macao predict. |
| 關鍵詞 Keywords |
互聯網搜索熱度，長短期記憶網絡，卷積長短期記憶神經網絡，旅遊需求，機器學習 Internet search popularity, LSTM (Long Short-Term Memory Network), CNN-LSTM (Convolutional Long Short-Term Memory Neural Network), Travel demand, Machine learning |
| 下載 Download |
Link |