Penerapan Automatic Clustering pada Fuzzy Time series pada Data Wisatawan Mancanegara Kalimantan Timur

Abstrak

Tourism played a significant role in national foreign exchange earnings and Regional Original Revenue (PAD), therefore accurate statistical analysis was needed as a preventive measure. Forecasting was one of the accurate statistical analyses that could assist the government in determining more effective policies in the future. The method used was the Automatic Clustering Fuzzy Logical Relationship (ACFLR) for time series data. Automatic Clustering was used to determine the length of data intervals, while the Fuzzy Logical Relationship was used to obtain forecasting results. The research objective was to forecast the number of foreign tourist visits for the next period using data on the number of foreign tourist visits to East Kalimantan from January 2023 to March 2024. The accuracy of the forecast was measured using the Mean Absolute Percentage Error (MAPE). The research findings indicated that the forecast for April 2024 was 261 visits with a MAPE value of 7.72%, indicating a very good level of accuracy. The conclusion of this research showed that the ACFLR method was effective in forecasting the number of foreign tourists, thus it could be used as a decision-making tool by local governments.