Bitte verwenden Sie diesen Link, um diese Publikation zu zitieren, oder auf sie als Internetquelle zu verweisen: https://hdl.handle.net/10419/305389 
Erscheinungsjahr: 
2024
Schriftenreihe/Nr.: 
CBM Working Papers No. WP/06/2024
Verlag: 
Central Bank of Malta, Valletta
Zusammenfassung: 
This paper presents a Machine Learning approach adopted at the Statistics Department of the Central Bank of Malta to detect outliers in the Maltese Pension Schemes and Insurance datasets, which are collected by the Bank, at micro-level. The motive behind this study is to develop an outlier detection model which can detect outliers in both short and long time series. The model used to detect these anomalies in our data is the unsupervised Machine Learning Algorithm known as Isolation Forests. The model is used to capture large discrepancies between submissions for balance sheet data. The algorithm is implemented using KNIME Analytics Platform, with the use of the KNIME/Python integration feature. We describe our results using various tables and graphs, discuss the effectiveness and efficiency of the tool, and conclude by explaining some limitations and further improvements.
Dokumentart: 
Working Paper

Datei(en):
Datei
Größe
796.94 kB





Publikationen in EconStor sind urheberrechtlich geschützt.