Innovation science and technologiy Том 2 № 5 (2026)
COMPARATIVE ANALYSIS AND EXPERIMENTAL EVALUATION OF ALGORITHMS FOR RECOVERING MISSING (NAN) VALUES IN INFORMATION SYSTEM DATA
Yarmatov, Sherzodjon, Orifov, Oxunjon
Аннотация
This article investigates the problem of identifying and recovering missing (NaN) values ininformation system data and examines their influence on analytical results. During the research process, artificialmissing values with different proportions were generated based on a complete dataset and subsequentlyrestored using statistical and machine-learning-based imputation methods. The effectiveness of each algorithmwas evaluated using error metrics obtained through comparison with the original ground-truth values. Theobtained results made it possible to determine the efficiency of different methods depending on the structure ofthe data and to establish a methodological basis for selecting optimal approaches in the intelligent analysis ofinformation system data. The findings of the study contribute positively to improving data quality and enhancingthe reliability of analytical processes
NaN values, missing data, imputation algorithms, KNN imputation, MICE method, data analysis, machine learning, information systems.
Источник метаданных: OAI-PMH архив журнала · Sindex не хранит полный текст, а даёт ссылку на источник.