EQNNs outperform classical counterparts in binary classification tasks with fewer parameters and smaller training datasets.