The main objective in the cryptanalysis of any unknown cryptographic algorithm is to determine the algorithm used for encryption before performing any targeted analysis. In this paper, the authors propose a novel method based on deep learning for identifying correlation between the initial keystream bytes of various stream ciphers. The primary objective of this method is to distinguish a stream cipher based on prediction accuracy of its keystream bytes at precise locations. This analysis can be done in a known-plaintext attack scenario. The authors validate their approach by analyzing six stream ciphers using two deep learning models and by providing a comprehensive discussion of the results. Their method successfully finds correlation in the keystream bytes of two stream ciphers, specifically RC4 and NGG. The graph showing the prediction accuracy values of keystream bytes at various positions, enables the differentiation of these two stream ciphers from the others. Additionally, the authors compare the accuracy of two deep learning models, Long Short-Term Memory (LSTM) and Multilayer Perceptron (MLP) which are used in this analysis.