The aim of the paper is to detect and identify diagnostic symptoms based on parametric modeling with the use of system identification methods in the scope of monitoring techniques intended for rotating machinery under transient operational conditions. The development effort should focus on early warning methods in order to increase detectability and performance of machines operation. The diagnostic symptoms may prove to be a powerful tool for the decision support systems based on easier interpretable parameters of a parametric model. The paper discusses the experimental results obtained with the use of a laboratory test rig as well as data from conducted numerical simulations.