DIGITAL ANALYSIS AND FORECASTING OF BIG DATA OF THE EDUCATIONAL PROCESS BASED ON THE LOGINOM PLATFORM
Back to articles of current journal| PDF: |  |
Author(s): | Kumratova A. M., Paraskevov A. V., |
| Number of journal: | 1(70) | Date: | March 2025 |
| Annotation: | Big data surrounds users regardless of their desire. Enterprises of any kind of activity often generate them in a huge volume. At the same time, on average, only one in three tries to process them. With a favorable combination of events, the figure reaches 35.5 %. And this is only due to the areas of information technology, design bureaus and sales. The authors aimed to con- duct a study of the admission campaign of a higher educational institution based on the generated data. The data had to be brought as close as possible to the real ones due to the correspondence of the competition indicators, the passing score, the percentage of applicants from cities and districts, and other indicators. The results and the entire course of the study were based on the data generated, but despite this, the results are of genuine scientific and practical interest. They demonstrate a methodology for evaluating performance indicators, regardless of the direction, forms of own- ership and seasonality of work. Based on the results of the study, a representative data set was generated; data selection, encoding and normalization operations were performed. The data obtained are as close as possible to the real ones due to the use of real indi- cators of the passing score and the number of applications sub- mitted. The real data was taken from open sources and is deper- sonalized. According to the results of the study, combinations of secondary factors that affect the applicant’s admission to higher education were identified. The key characteristics of the target audience of the educational institution were determined. A neural network was trained and the ability to predict the results of the next admission campaigns was implemented. Big data analysis methods and machine learning often play a crucial role in determining the quasi-optimal parameters of the functioning of both individual pro- cesses and industries as a whole. At the same time, it is important to observe the principle of variability of approaches. Clustering intermediate results or initial groups helps to achieve significant success. This defines a new look at the existing system, regardless of the industry. | ||
| Keywords: | higher education, big data, assessment meth- odology, data analysis, data set, dependencies, machine learn- ing, dependency research, parameter setting, analytics, neural networks |
||
| For citation: | Paraskevov A. V., Kumratova A. M. Digital analysis and forecasting of big data of the educational process based on the Loginom platform. Biznes. Obrazovanie. Pravo = Business. Education. Law. 2025;1(70):58—65. DOI: 10.25683/VOLBI.2025.70.1203. |
||