A kaggle data mining project, refer to https://www.kaggle.com/c/airbnb-recruiting-new-user-bookings
Zhao Hengrui(G1801739D) Luo Shuang (G1702502A) Wang Sixin(G1801764B)
MIT LICENSE Copyright@Nanyang Technological University
In the 21st century, the development of internet provides a window for people to know more about this colorful world, which drives them to desire to go to different cities they like. Compared with the past, they’re willing to spend money on travelling to enjoy their life with the improvement of people’s living standards. This leads to rising demand for travelling house booking. In order to solve this problem, Airbnb put up with an innovative idea --- providing a platform for travelers to book empty rooms in hosts’ houses. This attracts a lot of users because it can not only reduce the cost of travelers, but also make full use of empty space of hosts. Now, there are more than 150 billion travelers and 640,000 hosts in Airbnb. Their serve region covers 190 countries and 65000 cities (Smith, 2018). With more and more users, Airbnb hosted a recruitment competition, which requires to predict which country or city that the users will like to choose for their first booking by machine learning methods. Our project aims at using different algorithms (Logistic Regression, Tree, SVM, XGBoost) to develop models to identify users’ behavior patterns. By comparing those results, we will choose the best one for predicting new users’ first booking destination of Airbnb. This project can help to predict the most desirable travel destination for guests, followed by a lot of benefits. First of all, it can assist Airbnb to customize personal travel plan for traveler. What’s more, by providing the most desirable destination, it will motivate them to book within less time, which can increase booking rate. Beyond that, this kind of service can establish a good reputation of Airbnb so that they can attract more users. As for travelers, this can reduce their time for searching and deciding their destinations. As for hosts, this can help them to make strategies of location choice.
[1] Zabokrtsky, Z. (2015). Feature engineering in machine learning. Institute of Formal and Applied Linguistics, Charles University in Prague.
[2] Zheng, H., Yuan, J., & Chen, L. (2017). Short-term load forecasting using EMD-LSTM neural networks with a Xgboost algorithm for feature importance evaluation. Energies, 10(8), 1168.
[3] Chen, T., & Guestrin, C. (2016, August). Xgboost: A scalable tree boosting system. In Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining (pp. 785- 794). ACM.
[4] Smith, C. (2018). 100 Amazing Airbnb Statistics. Retrieved from https://expandedramblings.com/index.php/airbnb-statistics/
[5] Cortes, C., & Vapnik, V. (1995). Support-vector networks. Machine learning, 20(3), 273-297.
[6] Ulfsson, H. (2017). Predicting Airbnb user's desired travel destinations.
Fork or reference, please indicate the source @ Henry Zhao . Thx