An Improved Hybrid Algorithm for Social Network Analysis

The increasing trend of on-line social networks in different fields of knowledge, social network observations recently became the center of research. This paper proposes and implements an improved hybrid technique. The data set is taken from SNA. We analyze the Performance of Classification (PoC) Accuracy in Test Data. We applied our proposed algorithm on social network dataset and found average testing error. We compared it with back-propagation and found that our hybrid average testing error is significant and less than back-propagation.