Adaptive Artificial Immune Network Classification

The purpose of Artificial Immune System is to extract special information processing mechanisms contained in biological immune system, and then to study and design the corresponding models and algorithms, which can be used to solve many kinds of complex problems. As the hotspot of intelligent computation research field, Artificial Immune System has been applied in many fields such as information security, model recognition, intelligent optimization, automation control, data mining, where its strong ability of information processing and problem practicable solution is embodied. Artificial Immune Network is a successful model in Artificial Immune System, which has been widely used in data mining and machine learning. Based on Jerne’s immune network theory, a novel classification algorithm is proposed in this dissertation. Then this dissertation studies the comprehensibility of the immune network and two application algorithms are proposed:Intrusion Detection System based on Artificial Immune Network and Handwritten Digits Recognition based on Artificial Immune Network. The main research work of this thesis can be summarized as follows:1. Based on Artificial Immune Network Theory, a new Artificial Immune Network Classification algorithm is proposed. It adapts the parallel presenting and learning mechanisms. Further, an adaptive learning coefficient and a new antibody neighborhood are introduced. Based on immune clone operator and death operator, the network can accurately determine its structure. The results of experiments indicate that the new algorithm achieves better accuracy and data reduction.2. A novel algorithm of intrusion detection system based on adaptive artificial immune network is proposed. Combined with the similarity of Intrusion Detection System (IDS) and artificial immune system, artificial immune network classification algorithm is applied to Intrusion Detection. Besides, an application model of IDS based on Artificial Immune Network Classification is built.3. An improved algorithm of Handwritten Digits Recognition based on adaptive artificial immune network is presented. It explores the feasibility and effectiveness of artificial immune network classification applied in Handwritten Digits Recognition. The results of experiments show that the algorithm is viable and has good performance.