The Concept of Detecting and Classifying Anomalies in Large Data Sets on a Basis of Information Granules

Anomaly (outlier) detection is one of the most important problems of modern data analysis. Anomalies can be the results of database users' mistakes, operational errors or just missing values. The problem is important because of fast growth of the large data sets. Therefore, we present the initial results of work on a Granular Computing approach to data imputation and missing data analysis. Our proposal brings intuitive and interpretable solutions. Finally, in a series of experiments, we demonstrate its effectiveness for a large dataset in the area of transport.