No abstract

In data mining two important tasks involved are classification and clustering. In general, in classification the classifier assigns a class label from a set of predefined classes to a new input object. Whereas, given a set of objects, clustering creates different groups of these objects using some similarity measure. In the context of machine learning, classification is supervised learning and clustering is unsupervised learning. There are different approaches used for classification and clustering. In recent past many fuzzy neural networks have been proposed which can be employed for classification and clustering. Unlike other techniques, the fuzzy neural networks are quickly trainable, suitable for online training, provides soft decision, and capable of constructing nonlinear decision boundaries. All these benefits make them suitable for difficult real world problems involving classification and clustering. This paper provides review on recent fuzzy neural learning algorithms and mainly focusing on pattern/object classification and clustering.

Reliable information security mechanisms are required in the today"s era of cyber theft. Traditional Cryptography is a powerful mechanism to achieve information security. Current cryptographic algorithms (e.g., Advanced Encryption Standard (AES), Data Encryption Standard (DES), RSA) have a very high proven security but they suffer from the key management problem. Human identification is also a challenging field. That is why if we can make a blend of cryptography and biometric then it becomes a upcoming security tool. Using unique biometric identity of a person the keys for cryptosystem can be made secure. Iris is one of the proven and accurate means to identify person and it does not change throughout life of a person. This paper presents a biometric solution to cryptographic key management problem using iris based fuzzy vault.

No abstract

No abstract