On the diversity and similarity of mathematical models in science