Didyk O. Methods and algorithms of of control process models building based on retrospective data

Object of research: empirical knowledge accumulation tools of decision support systems. Purpose of research: increase of decision support systems functioning effectiveness by developing of new methods and algorithms of empirical knowledge acquisition by processing of retrospective data, accumulated during decision support systems functioning. Methods of rough set theory, expert evaluation, machine learning, databases theory and relational algebra was used. Methods of knowledge discovery in databases were amended with the purpose of control process models building of different objects. Method of control process graph models building was developed, that were improved by insertion of weight coefficients and inference functions of process operation execution conditions. Using of developed methods and algorithms allow to increase decision support systems functioning effectiveness by decision taking time reducing and quality increasing, that is leading to properties improvement of control process on the who le. Applied program system has inculcated in Department on extreme situations of the Kherson regional state administration.