Thailand is a country that attracts tourists from around the world. Thailand ranks 36th out of 117 countries globally and 3rd in ASEAN after Singapore and Indonesia in terms of tourism competitiveness (World Economic Forum, 2021). The Tourism Authority of Thailand (TAT) has organized an integration meeting for its 2024 Action Plan (Tourism Authority of Thailand Action Plan 2024: TATAP 2024) to set marketing promotion directions aiming to drive international tourist revenue to rank among the world's top 5 (Thanawan Winaisathian, 2023). As a result, competition in the hotel market has increased more than before. Not only does the hotel location affect tourist attraction, but physical characteristics also help attract tourists to choose hotels (Soifer, Choi & Lee, 2020). Currently, the internet is another variable influencing human life. People can share various information online, including new channels that help hotels reach tourists, creating more choices and increased competition. This has resulted in investment problems in both new hotel construction and hotel renovation due to higher management costs and economic uncertainty. Hotel business revenue primarily comes from room sales, accounting for 65-70% of total revenue. Therefore, understanding consumer willingness to pay will positively affect accommodation pricing strategies and allow businesses to adjust prices to maximize revenue while aligning with consumer needs in a subtle and efficient way (Perez & Lopez-Ospina, 2022). Due to the large amount of online data, computer systems have been developed that can learn through human training to help collect and analyze this data, called Machine Learning techniques, which work well in collecting and analyzing large amounts of data better than humans.

JEL Codes: L83, C45, D12

Keywords: Machine Learning, Deep Learning, 4-5 Star Hotel, Hotel Attribute, Willingness to pay