I. Introduction
The Electroencephalogram (EEG) signals are recorded over the scalp using multiple numbers of electrodes to obtain brain electrical signals. In existing methods EEG signals are recorded by extracting the patterns of EEG during sleeping condition and analysed using discrete wavelet transform (DWT) [1] & [2]. Statistical parameters like variance, SD, Energy etc are used to calculate sub band coefficients to identify the sleep disorders (T.V.K.H Rao et al and Omer Turk et al.,). The disadvantage is the inability to detect the different sleep disorders and less efficient and natural [3] & [4]. The proposed method data sets were well trained and analysed by using developed algorithms to determine the level of success of neural networks [4] & [5]. This project proposed to detect the sleep disorder from EEG signals by using deep learning neural networks [6] & [7].