Internet of Things (IoT) expands the internet connectivity into billions of IoT devices to collect, access, and share information. IoT devices heterogeneously generate huge amounts of stream data. Missing and insufficient values in the data streams of the IoT devices remain a challenging problem due to various reasons such as noise, collision, unstable network communication, equipment failure, and manual system closure. Compared to other learning models, deep learning possesses a high ability to handle and impute the missing values in the data stream. The deep learning model utilizes an imputation mechanism and deep neural network to estimate missing values and predict future values. Deep neural network-based stream data imputation for IoT application accurately predict the missing values.