Classification of yeast data plays an important role in the formation of medicines and in various chemical components. If the type of yeast can be recognized at the primary stage based on the initial characteristics of it, a lot of technical procedure can be avoided in the preparation of chemical and medical products. In this paper, the performance two classifying methodologies namely artificial neural network and fuzzy rule base has been compared, for the classification of proteins. The objective of this work is to classify the protein using the selected classifying methodology into their respective cellular localization sites based on their amino acid sequences. The yeast dataset has been chosen from UCI machine learning repository which has been used for this purpose. The results have shown that the classification using artificial neural network gives better prediction than that of fuzzy rule base on the basis of average error.

Security issues have been emphasized in MANET due to its vulnerability to unauthorised access and unshielded broadcasting nature of communication. In this paper we present a trust based AODV for MANET. The trust takes into account the eligible neighbours based on reliability, residual energy, and speed. Thus our algorithm provides a reliable, energy efficient routing technique. The multi-criteria trust values are calculated using fuzzy-logic. This algorithm is capable of putting aside the selfish nodes. As only trusted neighbours are selected for packet delivery, energy consumption also diminishes because the transmitting node does not need to deliver packets to the untrusted neighbours. Less number of transmissions renders low energy consumption. Absence of selfish nodes in the selected neighbours at every hop provides better packet delivery and hence better throughput.