A new approach for prediction of graphene based ISFET using regression tree and neural network

Abstract In this work, ion sensitive field effect transistor (ISFET) which is a device sensitive to the ions in a solution is employed. It is shown that under a fixed bias configuration, the voltage change causes a subsequent change in the surface potential of graphene thin film, which induces a detectable current change in the conducting channel between drain and source electrodes. Thus the transduction from an analog signal as an ion concentration(K+) changes to an electrical signal as current change can be achieved. For prediction purpose, the regression tree algorithm and artificial neural network (ANN) have been employed to predict the I-V characteristic, however ANN outperforms the regression tree approach and gives more accurate results.