On the Development of Smart-Watch Detection Using C 5.0 and S 5.24.1 Algorithms

Abstract

Smart-Watch as an emerging IoT solution in order to recognize human activities with respect to payment transactions. To implement this, a web application was developed specifically for payment transactions. The Smart-Watch is embedded within the transaction platform to monitor transactions on the IoT platform, and detect when a transaction is suspected to be a fraudulent. For the purpose of classifying transactions, we considered the C5.0 algorithm and S5.24.1 to present reliability of the Smart-Watch. The result from the software developed was tested using Confusion Matrix with a dataset of 100 transactions. The result obtained shows almost 93% of accuracy in detecting fraud using Smart-Watch. The Smart-Watch system has a high accuracy, and it will be a reliable system and also a profitable investment if put into action in the business world.

Keywords

References