Unemployment rates can greatly impact a country's economy. Predicting these rates accurately is crucial for economic planning. Traditional methods for forecasting unemployment rates have limitations due to the complex nature of the data. A new hybrid model combining linear and nonlinear approaches has been developed to improve predictions. This model considers the asymmetry of unemployment rates and has been tested on data from various countries. The results show that this hybrid model outperforms traditional methods, ensuring stable and reliable forecasts over time.