Application of ANFIS and Sobol approaches for optimization and sensitivity assessment of triglyceride biosorption from blood serum

Adaptive neuro-fuzzy inference system (ANFIS) was applied to simulate batch adsorption of triglyceride (TG) from the human blood serum using cinnamon powder, which has appeared as a potential serum-contact biosorbent, in our previous work. Experimental results were used to train and evaluate the ANFIS model. Serum temperature, contact time, stirring rate, cinnamon dose and particle size were considered as the model inputs and TG removal percentage was chosen as the model response. ANFIS model was trained with 75% of the available data while, 25% of the remaining data was used to verify the validity of the data-based model. Sobol sensitivity analysis results indicated that the adsorbent dose with %71 and particle size of the cinnamon with %15 share impact were the most affective variables on the adsorption performance. The specific surface area and the reactive adsorption sites density were obtained to be the most important characteristics of the adsorbents. Results of this study confirmed the advantages of ANFIS and Sobol approaches for data-based optimization of bioprocesses.