ASSESSMENT OF REGRESSION MODEL ADEQUACY BASED ON ERRORS IN EXPERIMENTAL MEASUREMENTS
Antonov Vladimir Aleksandrovich
The Institute of Mining of Ural Branch of the Russian Academy of Sciences
Abstract. The paper represents methodical techniques in the assessment of regression model adequacy, which describes the regularity of the change of a dependent variable taking into account its single and multiple dimensions separated conventionally on the basis of the proposed criterion. The realization of multiple measurements effect, as it is shown in the example, can improve the accuracy of the desired regularity detection.
Key words and phrases: экспериментальные измерения, закономерность, случайные отклонения, регрессия, модель, коэффициент детерминации, experimental measurements, regularity, random deviations, regression, model, determination coefficient
Open the whole article in PDF format. Free PDF-files viewer can be downloaded .
References:
Antonov V. A. o dostovernosti funktsional'no-faktornykh uravnenii regressii s samoopredelyayushchimisya parametrami // Glubinnoe stroenie, geodinamika, teplovoe pole Zemli, interpretatsiya geofizicheskikh polei: shestye nauchnye chteniya pamyati Yu. P. Bulashevicha: materialy konferentsii. Ekaterinburg: IGf UrO RAN, 2011. S. 17-20.
Antonov V. A., Yakovlev M. V. Otobrazhenie gorno-tekhnologicheskikh zakonomernostei funktsional'no-faktornymi uravneniyami nelineinoi regressii // Problemy nedropol'zovaniya: gornyi informatsionno-analiticheskii byulleten'. 2011. Otdel'nyi vypusk. S. 571-588.
Русский
