Evolving Fuzzy Neural Network Based on Uni-nullneuron to Identify Auction Fraud

The increase in transactions on the Internet related to the purchase of products or services can provide facilities for the parties involved in these acquisitions, but they also generate uncertainties and possibilities of attacks that can originate from fraud. This work seeks to explore and extract knowledge of auction fraud by using an evolving fuzzy neural network model based on n-uninorms. This new model uses a fuzzification technique based on Typicality and Eccentricity Data Analysis operators and a parallel processor for stream samples. To test the model in solving auction fraud problems, state-of-the-art neuro-fuzzy models were used to compare a public dataset on the topic. The results of the model proposed in this paper were superior to the other models evaluated (close to 96% accuracy) in the test, and the fuzzy rules demonstrate the model’s ability to extract knowledge.