Population identification strategies for counterfeit coin detection.

Area of Science:

• Forensic Science
• Numismatics
• Statistical Analysis

Background:

• Globalized manufacturing and advanced technology increase counterfeit coin production.
• Coin authentication is crucial due to rising fraud and economic incentives.
• Product-inherent features and reverse-quality-engineering aid in detecting fakes.

Purpose of the Study:

• To explore the use of finite mixture models for coin authentication.
• To compare coin weight distributions between known authentic and questioned coin sets.
• To assess the effectiveness of statistical methods in identifying fraudulent coinage.

Main Methods:

• Utilizing finite mixture models to analyze coin properties.
• Comparing the weight distribution of a 'Questioned' coin set against an 'Example' set of authentic coins.
• Applying statistical analysis to population measurements for authentication.

Main Results:

• Just over half of the 'Questioned' coins matched the weight distribution of authentic 'Example' coins.
• Nearly half of the 'Questioned' coins did not align with the authentic coin population's weight distribution.
• The findings suggest a significant presence of non-authentic coins within the 'Questioned' set.

Conclusions:

• Finite mixture models offer a viable statistical approach for coin authentication.
• Analyzing population properties, like weight, can effectively differentiate authentic from fraudulent coins.
• Combining multiple property analyses would strengthen the validation of coin authenticity findings.