Abstract
Here we report the Artificial Neural Network (ANN) along with theoretical model for the gain of the irradiated Thulium Doped Fiber Amplifier (TDFA). TDFA has been irradiated by fast neutrons up to 750Gy using Am-241/Be-9 source. ANN based on experimental data has been designed for irradiated TDFA. It consists of two inputs (wavelength and dose), one output (gain), two hidden layers and 50 neurons. The trained ANN model is successfully adapted to obtain the gain of TDFA at 1000, 2000 and 3000Gy. The TDFA gain reduced 22% when it is exposed to 3000Gy.