A cognitive radio system participates in a continuous process, the ‘‘cognition cycle”, during which it adjusts its operating parameters, observes the results and, eventually takes actions, that is to say, decides to operate in a specific radio configuration (i.e., radio access technology, carrier frequency, modulation type, etc.) expecting to move the radio toward some optimized operational state. in such a process, learning mechanisms that are capable of exploiting measurements sensed from the environment, gathered experience and stored knowledge, are judged as rather beneficial for guiding decisions and actions. framed within this statement, this report introduces some learning schemes that are based on deep learning techniques and can be used for predicting the capabilities (e.g. data rate) that can be achieved by a specific radio configuration. it also elaborates on different modulation identification methods for classification of cellular networks pertaining to automatic modulation classification (amc). with the help of matlab and neural network toolbox, modulation classification of am, bpsk, gsm signals, cdma2000 and data rate prediction using recurrent neural network (rnn) is simulated. this report also opens up path for classification of network signals using various deep learning networks with the help of feature based algorithms and ultimately helping the cr to efficiently predict the white spaces in the licensed spectrum for efficient spectrum sharing.