Automatic Emotion Recognition (AER) from speech is one of the most important sub domains in affective computing. Recent technological advances have enabled human users to interact with computers in ways previously unimaginable. Beyond the confines of the keyboard and mouse, new modalities for human-computer interaction such as voice, gesture, and force-feedback are emerging. This paper explores the Linear Prediction Coefficients (LPC) of speech signal for characterizing the basic emotions from speech. The emotions used in this study are sad, anger, happy, disgust, fear, and boredom. For capturing the emotion specific information from these higher order relations, neural network (NN) is used. The decrease in the error during training phase of the NN's and the emotion recognition performance of the models, demonstrate that the excitation source component of speech contains emotion-specific information and is indeed being captured by the NN.