This study investigates different modeling approaches and compares for drying of baker's yeast in a fluidized bed dryer. Four modeling concepts were investigated: modeling based on the mass and energy balance, modeling based on diffusion mechanism in the granule, modeling based on recurrent Artificial Neural Network (ANN) and modeling based on Adaptive Neural Network Fuzzy Inference System (ANFIS). Dry matter of product, product temperature and product quality were predicted using these model structures. To evaluate performances of the modeling structures, industrial scale drying process data were used. The vast numbers of industrial data (570 data sets) used for training and testing of the models were collected from a production-scale baker's yeast drying process.