Sleep is an essential process for the survival of animals. However, its phenomenon is poorly understood. To understand the phenomenon of sleep, the analysis should be made from the activities of a large number of cortical neurons. Calcium imaging is a recently developed technique that can record a large number of neurons simultaneously, however, it has a disadvantage of low time resolution. In this paper, we aim to discover phenomena which characterize sleep/wake states from calcium imaging data. We made an assumption that groups of neurons become active simultaneously and the neuronal activities of groups differ between sleep and wake states. We used non-negative matrix factorization (NMF) to identify those groups and their neuronal activities in time from calcium imaging data. NMF was used because neural activity can be expressed by the sum of individual neuronal activity and fluorescence intensity data are always positive values. We found that there are certain groups of neurons that behave differently between sleep and wake states.