The ability to have prediction about the future helps us in taking measured steps in the present. In this work we have predicted data well ahead in time using a novel approach based on Artificial Neural Networks (ANNs). This paper focuses on predicting both non-linear and linear data. For linear data we design a system which can predict the Grade Point Average (GPA) of a student based on his past track record. Neural network learns using the Back Propagation Algorithm. After learning from training sets, it accurately predicts the GPA of the student. Instead of using the conventional statistical methods, we learn the pattern using artificial neural network. Similarly for non-linear data we design a system which can predict the output current of a Semiconductor Diode, given an input voltage. The accuracy obtained for predicting GPA is around 97%, while for non-linear data the accuracy is 87.5%.

Full text: IJEIR_1231_Final.pdf