This document presents a comparative study on different pin geometries for tool profiles in friction stir welding using artificial neural networks. An experiment was conducted using AA6061 aluminum alloy plates welded with two different pin profiles (conical and triangular). Tensile strength test results were obtained for 27 welded joints under different welding parameters. Regression and artificial neural network models were developed to predict tensile strength based on the input parameters of tool rotation speed, welding speed, and axial force. The regression model results showed good agreement with experimental data. The artificial neural network model was found to be an effective method for predicting joint performance and identifying optimal welding conditions compared to regression analysis.