Выявление аномалии во временных рядах финансовой сферы

Abstract

Time series data is a valuable resource for analyzing various processes but can be challenging to work with due to the presence of anomalies. Autoencoder neural networks offer an effective solution for detecting these outliers by compressing and reconstructing data, identifying deviations. This approach is particularly useful in fields like finance, where anomaly detection help manage risks and improve decision-making.