With the ongoing development of global navigation satellite system (GNSS) technologies, user's dependence on these technologies has drastically risen. Ranging anywhere from industrial applications to agriculture or geodetic applications, the growing demand has stimulated the need for more sophisticated equipment with seamless accuracy. However, the signal received by a GNSS receiver has always faced the risk of contamination by many different sources of error, and this issue decreases the GNSS positioning accuracy. One of the most notable errors is the delay that occurs mainly in the troposphere layer, which mostly relies on the meteorological condition parameters. In this work, an artificial neural network (ANN)-based mitigation technique in order to eliminate troposphere delay is presented. Our ANN model trained using the surface meteorological parameters, namely, temperature, humidity, wind speed, maximum wind direction, and pressure with the number of satellites achieved by a GNSS receiver. The result showed that ANN model successfully improved the position accuracy level of the GNSS receiver in both horizontal (2D) and three dimensions (3D) at 43.77% and 19.19%, respectively.