An evaluation of the robustness of existing supervised machine learning approaches to the classification of emotions in speech