Many important real-world processes are modeled using systems of ordinary differential equations (ODEs) involving unknown parameters. The values of these parameters are usually inferred from experimental data. However, due to the structure of the model, there might be multiple parameter values that yield the same observed behavior even in the case of continuous noise-free data. It is important to detect such situations a priori, before collecting actual data. In this case, the only input is the model itself, so it is natural to tackle this question by methods of symbolic computation. We present new software SIAN (Structural Identifiability ANalyser) that solves this problem. Our software allows to tackle problems that could not be tackled before. It is written in Maple and available at https://github.com/pogudingleb/SIAN.
ASJC Scopus subject areas
- Computational Theory and Mathematics
- Computational Mathematics