The paper presents an entropy generation minimization study for a solar parabolic trough collector (PTC) operating with SiO2-water nanofluid using a genetic algorithm (GA), and artificial neural network (ANN). The characteristic variables of nanoparticle volumetric concentration (0.01 ≤ φ ≤ 0.05), mass flow rate (0.1≤ ṁ ≤ 1.1 kg/s), and inlet temperatures (350–550 K) are used to analyze the rate of entropy generated in the PTC. GA is used in optimizing the entropy generation rate for the specified parameters, while ANN is used for predicting and observing the behaviour of these parameters on the rate of entropy generation in the collector. The optimum ANN model is derived with one hidden layer of 18 neurons when training the input variables for the entropy generation predictions. The optimal mean square error used as a performance validation of the model is 0.02288 for training and 0.0282 for testing with an R2 value of 0.9999. The impact of the defined parameters on the entropy generation rate is presented in the results section. It is concluded that machine learning techniques can be an efficient tool for predicting the rate of entropy generation in a collector within the constraint of the defined parameters.