Similarity analysis

Similarity analysis is a mathematics-based database analysis method the essence of which is the objective comparison of a randomly selected composition of items (objects). The method is reasonable to use if there is a certain logical chain available and one logical factor (e.g. technical parameters) describes the other (e.g. price). The goal of the application is to involve subjective factors (e.g. weights, points) in the analyses to as limited an extent as possible.

The prerequisite of the analysis is the availability of an anonymous (i.e. one specified without line and column headings) O(bject)-A(ttribute)-M(atrix) (OAM).
As the result of the OAM analysis, the differences between the objects involved in the analysis become apparent and, from the differences, conclusions can be made as to which objects are in the balance position and which objects are under or overvalued.

It can be established as differences from or advantages compared to the database analysis methods currently applied at the market (e.g. decision trees, neural networks, etc.) that similarity analysis combines the advantages of other methods and guarantees more flexibility/higher efficiency in the analysis.

University partner: Szent István University, Gödöllő