Abstract
The purpose of this paper is to examine key methods for ensuring data quality for machine learning. This paper covers data cleaning and preprocessing processes, automated tools for data quality management, and ways to measure their suitability for model training. Data quality plays a key role in the success of machine learning models. Low-quality data can lead to distortions in models, lower forecast accuracy, and longer decision development time. The main aspects of data quality assurance are defined, which include data cleaning, omission processing, outlier management, normalization and standardization. Tools for automating monitoring and improving data quality, as well as metrics for evaluating their suitability for machine learning, are considered. The study confirms the importance of providing a high-quality data set for machine learning.