Wind energy is one of the sources which still in development in Brazil, however, it already represents 17% of the National Interconnected System. Due to the high level of uncertainty and fluctuations in wind speed, prediction of wind energy with high accuracy is a challenging task. In this context, this paper proposes a framework that combines Variational Mode Decomposition (VMD) based on Machine Learning algorithms to forecast the wind energy of a turbine in a wind farm at Parazinho city, Brazil, using a multi-step-ahead forecasting strategy. The forecasting models of the wind energy time series are Bayesian Regularization of Artificial Neural Networks, Cubist Regression, and Support Vector Regression. The performance of the proposed forecasting model, named VMD-CUBIST, was evaluated by using two performance metrics: symmetric mean absolute percentage error, and relative root mean square error. The VMD-CUBIST model outperforms the VMD, and single models in all forecasting horizons, with a performance improvement that ranges 3.00% - 83.30%. Indeed, VMD-CUBIST is an efficient and accurate model for wind energy forecasting. forecasting.