Along with the continuing growth of e-Commerce in North America, fraud and network intrusion cost e-Commerce companies an overwhelming lost each year. Fraud detection and network intrusion detection become more and more important to online e-Commerce business. However, data mining techniques in this domain are facing the challenges of large scale and high skewness of the data, missing and delayed labels, and the continuing change of patterns.;Experimental comparisons on both academic research data and practical real-world data for fraud detection and network intrusion detection demonstrate that the SICLN achieves high performance and outperforms traditional unsupervised clustering algorithms.;In this research, we develop two new clustering algorithms, the Improved Competitive Learning Network (ICLN) and the Supervised Improved Competitive Learning Network (SICLN), for the applications in the area of fraud detection and network intrusion detection. The ICLN is an unsupervised clustering algorithm applying new rules to the the Standard Competitive Learning Neural Network(SCLN). In the ICLN, network neurons are trained to represent the center of the data by a new reward-punishment update rule. The new update rule overcomes the instability of the SCLN. The SICLN is a supervised clustering algorithm further developed from the ICLN by introducing supervised mechanism. In the SICLN, the new supervised update rule utilizes the data labels to guide the training process to achieve a better clustering result. The SICLN can be applied to both labeled and unlabeled data and is highly tolerant to missing or delayed labels. Furthermore, the SICLN is completely independent from the initial number of clusters because it is able to reconstruct itself according to the labels of the cluster members.
展开▼
