<![CDATA[This research focuses on developing a prediction model for tourist visits for foreign destinations in the Asian region using Backpropagation algorithm architecture optimization. Tourism has become a crucial economic sector in Asia, and accurate tourist arrival predictions have a significant impact on decision making in this industry. The main approach used is the Backpropagation algorithm in the context of artificial neural networks. Although these algorithms have been successful in a variety of applications, optimizing Backpropagation architectures for tourism visit prediction remains a significant challenge. This research aims to improve model accuracy and performance by adjusting the Backpropagation algorithm architecture. Through careful optimization, this research seeks to overcome these complex dynamics and produce a model that can provide more accurate estimates of tourist visits. This research presents predictions of foreign tourist visits to Indonesia by optimizing the artificial neural network architecture using the Backpropagation algorithm. Focusing on visit data from various nationalities in the period 2018-2024, the test results highlight the performance variations between architectures in 2023 and 2024. Prediction results show that the 4-3-7-1 architecture obtains high test accuracy in 2023 (88%) , but will decrease in 2024 to 74%. The 4-5-1 architecture showed good consistency with test accuracy remaining high in both years (92%). These findings provide valuable insights for optimal architectural selection in predicting future tourist visits and identifying changing patterns of trends at the national level. However, it should be noted that these results are projective and may be influenced by external factors that may change, requiring ongoing evaluation to ensure model accuracy and responsiveness.]]>
Copyrights © 2024
