A four objective optimization on thermal and rheological properties of Aluminum nanoparticles dispersed in oil is studied. The four objectives are convective heat transfer, viscosity, thermal conductivity (TC), and specific heat capacity. Because of the high cost of experimental studies, response surface method and artificial neural network method were used as mathematical based and intelligent based methods respectively to predict four mentioned properties. Proposing a three-variable mathematical correlation as a function of temperature, solid volume fraction, and shear rate is another novelty of present work The best prediction accuracy by RSM, reported for viscosity prediction and the weakest prediction by this method was for TC prediction based on the amount of regression coefficients of 0.9914 and 0.9734, respectively. Also based on the mean square error as criteria, the best prediction accuracy by ANN reported for TC prediction with MSE of 4.28 x 10-7 and the worst prediction was for cp with MSE of 431163. (C) 2019 Elsevier B.V. All rights reserved.