Recommendation systems are aimed to suggest relevant items to the users based on user-s historical preferences. Traditional recommendation systems permit compact suggestions in the same domain appropriate to the user-s previous experience. It is necessary to make recommendations from different domains to satisfy user-s expectations. Domain adaptation aims to transfer the knowledge from one source domain to a different but related target domain. Domain adaptation supports a recommender system to produce multiple domain recommendations based on user preferences gained from related domains. Domain adaptation with deep learning models enables the system to automatically extract the predictive features from the distributed data that describe user preferences and complex user-item interactions. Deep learning-based domain adaptation for recommendation system exploits user preferences on a source domain to predict user preferences on a target domain and generate a useful recommendation that achieves users exceptions.