Title: Chain rules and subadditivities for Tsallis entropies

Abstract: A chain rule and a subadditivity for the information function of type $\beta$, which is one of the nonextensive (nonadditive) entropies, were derived by Z.Dar\'oczy. In this paper, introducing the different but natural conditional entropy based on Tsallis entropy, we firstly show the chain rules for the Tsallis type entropies which are also nonextensive (nonadditive) entropies, and the subadditivity of the Tsallis type entropies for the correlated two random variables in the case of $q \geq 1$ with the help of the generalized Shannon inequality. This idea enables us to further show the strong subadditivity of the Tsallis type entropies and then we can discuss on the entropy rate for the Tsallis entropies. Since the chain rules and the subadditivities imply that the conditioning reduces the Tsallis entropies, we are able to naturally define the Tsallis mutual entropy and also the Tsallis conditional mutual entropy. For these Tsallis mutual entropies, we again show the chain rules. Finally, so-called the data processing inequalities for the Tsallis mutual entropies are shown.