Recommender Systems are software techniques which can be used to filter out data from the volumes of data available online and provide recommendations to users in their area of interest. These techniques use information related to users and items in addition to the ratings given by users to various items or providing recommendations. In the last two decades, deep learning techniques have shown promising results in various areas of computer vision, video recognition, natural language processing etc. These techniques have been used for recommender systems in recent years and have shown improvement in performance. In this paper we propose a model, DeepAutoEnCF, that uses Denoising Autoencoder for predicting user ratings. It uses dropout for regularizing the model and adding noise to input for prediction of ratings. The model uses side information along with unique additional information for improving the performance.