Robust Statistical Outlier based Feature Selection Technique for Network Intrusion Detection
K. Nageswara Rao1, D. Rajyalakshmi2, T. Venkateswara Rao3

1K.Nageswara Rao, Research Scholar (part time), Gitam University, Visakhapatnam.
2Dr.D.Rajya Lakshmi, Professor and head, Dept of IT, Gitam University, Visakhapatnam.
3Prof.T.Venkateswara Rao, Professor, Dept of CSE, K.L University,Vijayawada.

Manuscript received on February 15, 2012. | Revised Manuscript received on February 20, 2012. | Manuscript published on March 05, 2012. | PP: 453-459 | Volume-2 Issue-1, March 2012. | Retrieval Number: A0477022112 /2012©BEIESP
Open Access | Ethics and Policies | Cite 
© The Authors. Published By: Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: A computational model could serve as a conventional engineering approach to uncover the biochemistry of the metabolic pathways. These would dynamically mimic the pathways in-silico. Flux Balance Analysis (FBA) is one such method wherein characterization of growth yields, bio-energy production, environmental conditions and robustness under knock out & knock down can be studied. We have built a comprehensive dynamic platform of integrated network for melanogenesis pathway containing 6 major reactions. Wherein detailed stoichiometric matrix of the pathway reactions is constructed followed by defining constrains and objective function. Subsequently, these are optimized using linear programming to give us resultant fluxes. Using this model, vulnerability of the enzymes in these pathways are studied; essentiality of participating enzymes are established and varied computational gene knock-out experiments which can decipher effect of inhibition on metabolic circuit are performed. Results of the simulations were in corroboration with published results and predictions were validated. However, this platform can enables us to make elaborate prediction in the known modeled domain and later with amalgamation of more modelled pathways into this network; a comprehensive virtual cell can be constructed.

Keywords: Normalization, Network security, data mining, intrusion detection, filtering.