Modelling atmospheric pressure through the hybridization of an ANFIS using IOWA and a snake optimizer

The atmospheric pressure plays a critical role in ecology because it serves as an essential indicator of environmental phenomena. Hence, reliable atmospheric pressure forecasts are crucial for the precise assessment of ecosystems. Therefore, the article investigates a novel framework of Adaptive Neuro-Fuzzy Inference System (ANFIS) for addressing a complex problem involving the hourly prediction of atmospheric pressure. In this study, ANFIS was hybridized with Snake Optimizer (SO), and dimensionality problems were addressed using Induced Ordered Weighted Average (IOWA). This framework is known as IOWA-ANFIS-SO. To accomplish hourly atmospheric pressure, the datasets featuring meteorological factors including air temperature, sea surface temperature, surge wind speed, wind speed, and wind direction were collected from weather buoy network stations in Ireland. The IOWA-ANFIS-SO model is assessed against the atmospheric dataset, using 70% of the data for training and 30% of the data for testing the model. Further, a comparison was made between the new hybrid IOWA-ANFIS-SO model and the more traditional IOWA-ANFIS at various alpha levels, utilizing the four statistical benchmarks: the RMSE, MAE, MAPE, and R². Among the models, IOWA-ANFIS-SO produced optimal results for hourly atmospheric pressure prediction with RMSE (0.4698), MAE (0.3593), MAPE (0.0003), and R² (0.9903) at alpha = 0. Further, the developed model was compared to IOWA-ANFIS and ANFIS-SO at various alpha levels, demonstrating that IOWA-ANFIS-SO outperformed them both. Finally, the results emphasized the importance of incorporating the IOWA and a snake optimizer as a means to augment the functionality of ANFIS. Therefore, this hybrid ANFIS mechanism might be advantageous for monitoring ecological systems with a wide array of input variables that are highly diverse and unpredictable.