Tourism planning
majid ghias; shohreh moradpour
Abstract
This research leverages machine learning techniques to predict promising tourist destinations in the western region of Isfahan Province. A random forest prediction model is employed to identify the factors influencing tourist attraction in these areas and subsequently predict potential tourist hotspots. ...
Read More
This research leverages machine learning techniques to predict promising tourist destinations in the western region of Isfahan Province. A random forest prediction model is employed to identify the factors influencing tourist attraction in these areas and subsequently predict potential tourist hotspots. The input data encompasses 34 sites, including springs, waterfalls, shrines, museums, historical structures, and more. Variables such as elevation, slope, aspect, vegetation cover, distance to rivers, distance to roads, temperature, and precipitation were selected as the most significant predictive variables based on the Variance Inflation Factor (VIF). Evaluation metrics reveal that the random forest model exhibits superior predictive performance, achieving the highest R² (0.91) and the lowest RMSE (1.06) and MAE (1.13). Slope, slope aspect, vegetation cover, and distance to rivers and roads were identified as the most critical predictors. Moreover, the spatial prediction map indicates that Chadgan and daran possess high potential for tourist reception.
