In the area of ecological data analysis there is an increasing demand for methods and tools based on novel approaches from machine learning and information theory that would complement classic statistical methods. This would significantly increase the number of tasks that can be addressed with data analysis and provide higher quality of the analysis results. Data mining, for example, uses machine learning methods that employ approaches from classical statistics as well as information theory. Machine learning tools have been successfully used for data analysis and learning of qualitative and quantitative models from the data. Models can be written in a human readable form (e.g., decision rules and trees, equations) or in a form that can only be used for predicting new examples (e.g., neural networks, support vector machines, etc.). For the purpose of the analysis of ecological data, decision trees are frequently the method of choice. Due to their hierarchical structure, models in the form of decision trees are easy to interpret and can be used to predict values of the target variable that can be simple or structured (e.g., a vector, a hierarchy, etc.). We present a pilot study of applying data mining techniques in order to analyze the forest in Slovenia and to develop scenarios of its future development that could be used in forest management. We have focused on the analysis of the time dynamics of wood stock changes and developed models that enable us to investigate the factors that influence the total wood stock in a forest, as well as to predict the wood stock in the future.