Authors:
Andriy Ryabko
1
;
Tetiana Vakaliuk
2
;
3
;
4
;
Oksana Zaika
1
;
Roman Kukharchuk
1
;
Viacheslav Osadchyi
5
and
Inesa Novitska
6
Affiliations:
1
Olexander Dovzhenko Glukhiv National Pedagogical University, 24 Kyievo-Moskovska Str., Glukhiv, 41400, Ukraine
;
2
Kryvyi Rih State Pedagogical University, 54 Gagarin Ave., Kryvyi Rih, 50086, Ukraine
;
3
Institute for Digitalisation of Education of the NAES of Ukraine, 9 M. Berlynskoho Str., Kyiv, 04060, Ukraine
;
4
Zhytomyr Polytechnic State University, 103 Chudnivska Str., Zhytomyr, 10005, Ukraine
;
5
Borys Grinchenko Kyiv University, 18/2 Bulvarno-Kudriavska Str., Kyiv, 04053, Ukraine
;
6
Zhytomyr Ivan Franko State University, 30 Velyka Berdychivska Str., Zhytomyr, 10002, Ukraine
Keyword(s):
Evaluation Criteria, Educational Program, Educational Activities, Prognostication, Rating, ANFIS, Artificial Neural Networks.
Abstract:
The article discusses a methodology for assessing the quality of educational programs and activities in higher education institutions using artificial intelligence tools such as the adaptive system of neuro-fuzzy inference (ANFIS) and an L-layer neural network. The purpose of the study was to address the problem of objectivity in self-assessment and identify potential problems and shortcomings in educational activities before the start of an accreditation examination. The study used student ratings on a four-level assessment scale as input data for the L-layer neural network, and the criteria for assessing the quality of the educational program as input variables for the ANFIS system. The hypothesis was that students with higher ratings of educational achievement would provide more objective assessments of the quality criteria of the educational program and activities. The results showed that the L-layer neural network made more accurate predictions than the ANFIS model. The article
suggests that this approach can provide higher education managers with qualitative forecasts to determine the quality of educational services and identify potential problems before the start of an accreditation examination. However, the study acknowledges the need for further research on larger data volumes to improve the predictive capabilities of the models.
(More)