An evaluation framework and a brief survey of decision tree tools
Nikola Vlahović · 2016
Over the last 50 years decision trees have been extensively used in numerous application areas by practitioners and academics alike. Decision trees are both a mathematical representation tool and a highly versatile graphic representation tool suited for modelling and visualisation of structured hierarchical concepts. Decision tree construction algorithms belong to supervised learning algorithms in the field of machine learning. Most prominent applications areas of decision trees are managerial sciences, knowledge engineering and data analysis. Variety of specific decision tree types is used: decision analysis trees, knowledge representation trees, classification and regression trees, decision forests, etc. Software tools for each of these types are available often borrowing from one decision tree concept to another. As a consequence, distinct features and capabilities of these tools make direct comparison difficult. Goal of this paper is to propose a framework for evaluation of available decision tree tools in relation to their foundation model. Brief survey of currently available decision tree tools will be given accordingly. Results may motivate more comprehensible studies of decision tree applications. Results may also motivate development of hybrid decision tree tools where sound combination of decision tree concepts can enhance functionalities and provide additional benefits to users.