在當今體育產業中，對於球團與聯盟而言，準確預測球賽的觀眾人數極為關鍵。這不僅牽涉到門票、場館營運與周邊商品的銷售，同時也影響媒體轉播策略和贊助商關係的管理。本研究運用日趨成熟的機器學習技術，融合比賽資訊、氣象條件以及新聞媒體報導等多元變數，並採用隨機森林模型、多變量迴歸分析和長短期記憶模型，旨在探究影響台灣籃球賽事觀眾人數的主要因素，並評估各模型的預測效果。
研究發現，台灣籃球賽事的觀眾人數受諸多因素影響，特別是主隊隊伍、比賽日程（工作日或週末）及新聞媒體的聲量。主隊隊伍這一因素涵蓋比賽地點、球員狀況、主場地區的繁榮程度等多重要素，是影響觀眾人數預測最為關鍵的變量。
在所採用的模型中，隨機森林模型表現最佳，其穩定性與準確度高，平均Adjusted R square 達0.820.82，預測誤差僅為371 人。相較之下，多變量迴歸模型在預測複雜問題如觀眾人數時表現不佳。至於長短期記憶模型，則由於資料量限制，未能充分展現深度學習的預測優勢。
綜上所述，本研究不僅為台灣職業籃球聯盟後續相關研究奠定了基礎，其結果亦可作為聯盟廣告策略、主場城市交通規劃等領域的重要參考，進而助力於提升台灣職業籃球聯盟的整體競賽環境品質。

In today's sports industry, accurately predicting the number of spectators at a game
is crucial for teams and leagues. This affects not only ticket sales, venue operations, and
merchandise sales, but also media broadcasting strategies and sponsorship relat ions. This
study uses advanced machine learning techniques, combining information about the game,
weather conditions, and media reports to explore factors affecting audience numbers at
basketball games in Taiwan, using models like Random Forest, Multivaria te Regression,
and Long Short Term Memory (LSTM).
The study finds that audience numbers in Taiwan's basketball games are influenced
by many factors, especially the home team, the date of the game (weekday or weekend),
and media coverage. The home team factor includes the game location, player conditions,
and prosperity of the home area, making it a key variable in predicting audience numbers.
Among the models used, the Random Forest model performs best with high stability
and accuracy, with an average Adjusted R square of 0.82 and an average prediction error
of only 371 people. In contrast, the Multivariate Regression model does not perform wel l
in predicting complex issues like audience numbers. The LSTM model, limited by the
amount of data, could not fully demonstrate the advantages of deep learning predictions.
In summary, this study not only lays the foundation for future research in the Taiwan
professional basketball league but also provides important references for the league's
advertising strategies, urban traffic planning in home cities, and thus contributes to
improving the overall quality of the competitive environment in Taiwan's professional
basketball league.

Andreff, W. (2001). The correlation between economic underdevelopment and sport. European Sport Management Quarterly, 1(4), 251–279. https://doi.org/10.1080/16184740108721902
Annamalai, B., Yoshida, M., Varshney, S., Pathak, A. A., & Venugopal, P. (2021). Social Media Content Strategy for Sport Clubs to Drive Fan Engagement. Journal of Retailing and Consumer Services, 62(62), 102648. https://doi.org/10.1016/j.jretconser.2021.102648
Chen, C.-K. (2012). Hierarchical linear relationship between the U.S. leisure and entertainment consumption. Technology in Society, 34(1), 44–54. https://doi.org/10.1016/j.techsoc.2011.12.003
Davies, L. E. (2002). Consumers’ expenditure on sport in the UK: increased spending or underestimation? Managing Leisure, 7(2), 83–102. https://doi.org/10.1080/13606710210137237
Eakins, J. (2015). An examination of the determinants of Irish household sports expenditures and the effects of the economic recession. European Sport Management Quarterly, 16(1), 86–105. https://doi.org/10.1080/16184742.2015.1067238
Griffith, D. A. (2010). An analytical perspective on sporting events attendance: The 2007–2008 US NCAA college bowl games. Applied Geography, 30(2), 203–209. https://doi.org/10.1016/j.apgeog.2009.01.005
Hirschle, J. (2014). Consumption as a Source of Social Change. Social Forces, 92(4), 1405–1433. https://doi.org/10.1093/sf/sou001
Karg, A., Nguyen, J., & McDonald, H. (2021). Understanding Season Ticket Holder Attendance Decisions. Journal of Sport Management, 1–15. https://doi.org/10.1123/jsm.2020-0284
Kim, T., Hong, J., & Kang, P. (2015). Box office forecasting using machine learning algorithms based on SNS data. International Journal of Forecasting.
King, B. E. (n.d.). Predicting National Basketball Association Game Attendance Using Random Forests : Journal of Computer Science and Information Technology. Jcsitnet.com. Retrieved December 11, 2023, from http://jcsitnet.com/vol-5-no-1-june-2017-abstract-1-jcsit
Lera-López, F., Ollo-López, A., & Rapún-Gárate, M. (2012). Sports spectatorship in Spain: attendance and consumption. European Sport Management Quarterly, 12(3), 265–289. https://doi.org/10.1080/16184742.2012.680897
Lera-López, F., & Rapún-Gárate, M. (2005). Sports Participation versus Consumer Expenditure on Sport: Different Determinants and Strategies in Sports Management. European Sport Management Quarterly, 5(2), 167–186. https://doi.org/10.1080/16184740500188656
Lera-López, F., & Rapún-Gárate, M. (2007). The Demand for Sport: Sport Consumption and Participation Models. Journal of Sport Management, 21(1), 103–122. https://doi.org/10.1123/jsm.21.1.103
Lizana, M., Carrasco, J.-A., & Tudela, A. (2019). Studying the relationship between activity participation, social networks, expenditures and travel behavior on leisure activities. Transportation. https://doi.org/10.1007/s11116-019-09980-y
Lövdal, S. S., Den Hartigh, R. J. R., & Azzopardi, G. (2021). Injury Prediction in Competitive Runners With Machine Learning. International Journal of Sports Physiology and Performance, 16(10), 1522–1531. https://doi.org/10.1123/ijspp.2020-0518
Mueller, S. Q. (2020). Pre- and within-season attendance forecasting in Major League Baseball: a random forest approach. Applied Economics.
Nisar, T. M., Prabhakar, G., & Patil, P. P. (2018). Sports clubs’ use of social media to increase spectator interest. International Journal of Information Management, 43, 188–195. https://doi.org/10.1016/j.ijinfomgt.2018.08.003
Pavlyshenko, B. (2019). Machine-Learning Models for Sales Time Series Forecasting. Data.
Popp, N., Jensen, J., & Jackson, R. (2017). Maximizing visitors at college football bowl games. International Journal of Event and Festival Management, 8(3), 261–273. https://doi.org/10.1108/ijefm-02-2017-0014
Şahin, M., & Erol, R. (2017). A Comparative Study of Neural Networks and ANFIS for Forecasting Attendance Rate of Soccer Games. Mathematical and Computational Applications, 22(4), 43. https://doi.org/10.3390/mca22040043
Şahin, M., & Uçar, M. (2020). Prediction of sports attendance: A comparative analysis. Proceedings of the Institution of Mechanical Engineers, Part P: Journal of Sports Engineering and Technology.
Skinner, J., & Smith, A. C. T. (2021). Introduction: sport and COVID-19: impacts and challenges for the future (Volume 1). European Sport Management Quarterly, 1(3), 1–10. https://doi.org/10.1080/16184742.2021.1925725
Thibaut, E., Vos, S., & Scheerder, J. (2014). Hurdles for sports consumption? The determining factors of household sports expenditures. Sport Management Review, 17(4), 444–454. https://doi.org/10.1016/j.smr.2013.12.001
Wakefield, K. (2016). Using Fan Passion to Predict Attendance, Media Consumption, and Social Media Behaviors. Journal of Sport Management, 30(3), 229–247. https://doi.org/10.1123/jsm.2015-0039
Wilhite, B., & Shank, J. (2009). In praise of sport: Promoting sport participation as a mechanism of health among persons with a disability. Disability and Health Journal, 2(3), 116–127. https://doi.org/10.1016/j.dhjo.2009.01.002
Wilkens, S. (2021). Sports prediction and betting models in the machine learning age: The case of tennis. Journal of Sports Analytics, 1–19. https://doi.org/10.3233/jsa-200463
Yoshizawa, Y., Kim, J., & Kuno, S. (2016). Effects of a Lifestyle-Based Physical Activity Intervention on Medical Expenditure in Japanese Adults: A Community-Based Retrospective Study. BioMed Research International, 2016, 1–6. https://doi.org/10.1155/2016/7530105