Refractive window drying is an efficient technique in order to retain the quality of the food product. So, in this study, thin layer drying behavior of raw banana (Musa sapiantum L.) was examined in refractive window dryer under the temperatures ranging from 70 to 90ºC. For describing the drying behavior three mathematical models were fitted. Among which Experimental model showed best fit with a R2 value of more than 0.99. In order to get better result artificial neural network (ANN) modeling was done. 3-25-1 was selected as the best ANN architecture. It was observed that ANN modeling (R2=0.998 ) gives better result than the mathematical modeling (R2=0.997). From this study, it was also summarized that accuracy is not the main reason due to which ANN models are superior over empirical relations. ANN models are better than the empirical relations due to their generality. A range of experiments can be described by the ANN modeling while empirical relations are valid for some specific experiments under controlled condition. Due to this reasons, ANN modeling can be implemented as one of the statistical modeling technique in order to describe the thin layer drying behavior of raw banana in refractive window dryer.