In this work, a method called Discrete Sine Transform is used in order to classify EEG signals of the brain and are used for the purpose of early autism detection. Autism is a complex behavioral disorder. The EEG signals are obtained by using electrodes that are attached to the brain. Each electrode measures the signals in different regions of the brain as the subject responds to different stimuli. The signals are processed by applying Discrete Sine Transform (DST) and then it is passed as an input to an artificial neural network. DST is used as it greatly simplifies the process and reduces the complexity. The computing tool of a neural network is used in order to objectively estimate whether a subject is suffering from autism. The network is trained with a particular dataset and upon applying a testing input the output was achieved. This study looked at EEG signals, an indirect measure of brain connectivity, and identified patterns that distinguished subjects at an increased risk for autism. © 2014 IEEE.