The article deals with the imitating modeling of the missing students’ knowledge testing results time series values reconstruction process with the time series singular spectrum analysis method. The analysis method, used in this work for research of time series with missing data, makes it possible to get exact results in assumptions strict enough, can be applied to real time series with missing data, leading in this case to the approximate results either. Statistical estimations of missing data filling algorithms with singular spec- trum analysis method efficiency for a concrete subject domain – training influences efficiency determination –have been received.
time series, Singular spectrum analysis, education quality, knowledge testing, training influences
"Analyz эffektyvnosty alhorytma vosstanovlenyia propushchennыkh znachenyi vremennoho riada rezultatov testyrovanyia znanyi" ,
Information Processing Systems,