Overcoming High Dimensionality: A Case of Consumer Anarchism

Abstract This study empirically tests ‘machine-learning’ (ML) regression models on a consumer anarchist dataset with high dimensionality. Currently there are not enough research papers addressing methodological problems caused by the high dimensionality of a dataset. Thus, this is the first of its kind to empirical research comparing newly evolving machine learning techniques in the context of consumer anarchism. The comparative results indicate that Random Forest (RF) models outperform the Earth (also known as MARS). Further, the results also revealed that consumer anarchists’ feelings and beliefs could be associated with the hated company’s socially irresponsible behaviors.