Use of multiple criteria decision aid methods in case of large amounts of data

Abstract

Cases of large amounts of data and high numbers of criteria for evaluation of socio–economic objects are rather frequent. Evaluation can comprise thousands of entries of data and dozens of different criteria. Quantitative methods of processing large amounts of data could be classified into two broad categories: statistical methods and multiple criteria decision aid (MCDA) methods. Statistical methods impose a number of rather strong limitations on data. In contrast, multiple criteria evaluation methods can deal with ill–defined problems and with multi–dimensional data. Results yielded by statistical methods can be comprised by specialists, while results yielded by the MCDA methods are specifically designed for decision–makers. The MCDA methods provide results in the form of ranking of alternatives by their preference to decision–makers of various backgrounds. Even if is a convenient way, it is not well–informative. In the paper, various techniques of choosing the most important criteria, of building a hierarchy of criteria, of retrieval of results of evaluation broadening usage of multiple criteria methods are proposed, making emphasis on cases with large amounts of data.