<![CDATA[Time series data represents measurements taken over a specific period and is often employed for forecasting purposes. The typical approach in forecasting involves the analysis of relationships among estimated variables.In this study, we apply Fuzzy Time Series (FTS) to water level data collected every 10 minutes at the Irish Achill Island Observation Station. The FTS, which is based on Fuzzy C-Means (FCM), is hybridized with the Cat and Mouse Based Optimizer (CMBO). This hybridization of FCM with the CMBO optimizer aims to address weaknesses inherent in FTS, particularly concerning the determination of interval lengths, with the ultimate goal of enhancing prediction accuracy.Before conducting forecasts, we execute the FCM-CMBO process to determine the optimal centroid used for defining interval lengths within the FTS framework. Our study utilizes a dataset comprising 52,562 data points, obtained from the official Kaggle website. Subsequently, we assess forecasting accuracy using the Mean Absolute Percent Error (MAPE), where a smaller percentage indicates superior performance. Our proposed methodology effectively mitigates the limitations associated with interval length determination and significantly improves forecasting accuracy. Specifically, the MAPE percentage for FTS-FCM before optimization is 20.180%, while that of FCM-CMBO is notably lower at 18.265%. These results highlight the superior performance of the FCM-CMBO hybrid approach, which achieves a forecasting accuracy of 81.735% when compared to actual data.]]>
