A recommender system is a process that investigates to predict user preferences. Common problems in classic recommendation systems are data sparsity and cold start problems. The emergence of cross-domain recommendations is a promising solution to address such a problem. A cross-domain recommendation system is a collective approach that manipulates feedback from multiple domains to improve the performance of the recommendation and also possesses the ability to address data sparsity problems. Deep learning models improve the recommender system by learning the hidden features of users and items from huge data and subsequently generate effective suggestions for the user. Incorporating deep neural network in cross-domain recommendation system utilizing more preference data and accurately map the dormant factors across domains. Deep neural network-based cross-domain recommendation collectively learns features of users and items from different domains and significantly outperforms other state-of-the-art recommendation approaches.