Four classification models of economic bankruptcy of farms based on panel data are presented in the article. The modelling techniques are linear discrimination function, pooled probit model, random effects probit model and artificial neural network. They give very similar quality of classification, so we can assume that the problem is linear.
Study presents a classification model for predicting the occurrence of mass damage in young forest plantations caused by insect pests on roots belonging to Melolontha genus. Logistic regression model was built on the basis of 10 taxation features describing 177 young stands. Habitat fertility, occurrence of Scots pine and European beech, and weed infestation of the habitat are the most significant features that influence mass occurrence of cockchafer larvae acting as insect pests on roots in plantations.