ABSTRACT
The main difference between supervised and unsupervised learning methods is the underlying model structure. In supervised learning, relationships between input (independent, predictor) and the target (dependent, response) variables are being established. However, in unsupervised learning, no variable is defined as a target or response variable. In fact, for most types of unsupervised learning, the targets are same as the inputs. All the variables are assumed to be influenced by a few hidden or latent factors in supervised learning. Because of this feature, it is better to explore large complex models with unsupervised learning than with supervised learning methods.
