-------------------------------------------------------------------------------------------
A novel multi-epitopic immune network model hybridized with neural theory and fuzzy concept 
-------------------------------------------------------------------------------------------

References and further reading may be available for this article. To view references and further reading you must purchase this article.


Hamid Izadinia, a, , , Fereshteh Sadeghia,  and Mohammad Mehdi Ebadzadeh1, a, 

aAmirkabir University of Technology (Tehran Polytechnic), Tehran, Iran

Received 5 May 2009;  revised 4 June 2009;  accepted 25 June 2009.  Available online 2 July 2009. 

Abstract
The natural immune system provides an effective defense mechanism against foreign substances via complex interactions among various cells and molecules. Jerne introduced the immune network theory to model the relation between immune cells and molecules. The immune system like the neural system is able to learn from experience. In this paper, a multi-epitopic immune network model is proposed. The proposed model is hybridized with 
Learning Vector Quantization (LVQ) and fuzzy set theory to present a new supervised learning method. The new method is called Hybrid Fuzzy Neuro-Immune Network based on Multi-Epitope approach (HFNINME). To evaluate the performance of the proposed method several experiments on benchmark classification problems are carried out and the results are compared with two prominent immune-based classifiers as well as several versions of the LVQ algorithm. The results of the experiments reveal that the proposed method yields a parsimonious classifier that can classify data more accurately and more efficiently.


Keywords: Immune network theory; Multi-epitope approach; Learning vector quantization; Fuzzy set; Artificial immune system

Article Outline
1. Introduction 
2. Related works 
     2.1. LVQ 
     2.2. AIRS 
     2.3. SRABNET
3. Immune network theory 
4. Proposed model 
5. Proposed method 
     5.1. Preprocessing 
     5.2. Coarse training phase 
     5.3. Fine training phase
6. Experiments and results 
7. Conclusion 
References


 

Fig. 1. Molecular interactions in Jernefs immune network model (Jerne, 1974a). 


Wikipedia  Learning vector Quantization
==================================================================================================
LVQ can be understood as a special case of an artificial neural network, more precisely, it applies a winner-take-all Hebbian learning-based approach. It is a precursor to Self-organizing maps (SOM) and related to Neural gas, and to the k-Nearest Neighbor algorithm (k-NN). LVQ was invented by Teuvo Kohonen.

The network has three layers: an input layer, a Kohonen classification layer, and a competitive output layer. The network is given by prototypes W=(w(i),...,w(n)). It changes the weights of the network in order to classify the data correctly. For each data point, the prototype (neuron) that is closest to it is determined (called the winner neuron). The weights of the connections to this neuron are then adapted, i.e. made closer if it correctly classifies the data point or made less similar if it incorrectly classifies it.

An advantage of LVQ is that it creates prototypes that are easy to interpret for experts in the field.

LVQ can be a source of great help in classifying text documents.