Development of an integrated approach to detecting anomalies in multidimensional data

The paper analyzes the impact of anomalous data on analytical models in high-dimensional and heterogeneous datasets. It investigates the use of metrics such as accuracy, completeness, and consistency to evaluate data quality in different scenarios. The study reveals that traditional approaches often fail to detect complex structural anomalies. The effect of parameters in the comprehensive data quality assessment on anomaly detection accuracy is studied. A new Integrated Approach for Anomaly Detection methodology is proposed, which combines a comprehensive data quality assessment with hybrid anomaly detection algorithms.The paper analyzes the impact of anomalous data on analytical models in high-dimensional and heterogeneous datasets. It investigates the use of metrics such as accuracy, completeness, and consistency to evaluate data quality in different scenarios. The study reveals that traditional approaches often fail to detect complex structural anomalies. The effect of parameters in the comprehensive data quality assessment on anomaly detection accuracy is studied. A new Integrated Approach for Anomaly Detection methodology is proposed, which combines a comprehensive data quality assessment with hybrid anomaly detection algorithms.