Abstract: The functioning of a currency is essential for the economy of a country, and banknotes are a major component of the Indian economy. Counterfeiting currency is an attempt to imitate a real currency with the intention of deception. Most of the methods used to detect counterfeit currency are based on hardware or image processing techniques, which are less efficient and time- consuming. As technology advances, the methods used to counterfeit currency have become more sophisticated, and the circulation of such notes has a significant impact on the economy. Therefore, the detection of counterfeit notes is of paramount importance. There are numerous commercial methods for detecting fake notes, however, they are not accessible to the general public. People who acquire counterfeit currency are often victims, and there is usually no government policy to reimburse them for the counterfeit notes that are confiscated. To design an automated system, it is necessary to develop an efficient algorithm that can accurately predict whether the banknotes are genuine or forged, as counterfeit notes are designed with great accuracy.

Keywords: Banknote Authentication, Fake-notes, Skewness, Logistic regression , XG boost, Decision tree classifier, Pre- processing.

| DOI: 10.17148/IARJSET.2024.11717