Neural network applications in physics

Our study describes many new opportunities for using neural networks in physics. We have mapped types of physics problems to analogous applications in other areas of science and engineering. While many applications are possible, little work can be found in the literature. Our specific example shows an interesting and useful application for predicting concentrations of radioactivity in the environment. Known levels of radioactivity, along with the values of other environmental variables, can be used to train a network for estimating subsequent levels. The accuracy of the neural network approach is better than other methods for specific monitoring locations. The possibility of finding generic patterns that can be used across different locations will be discussed