Scribd
Upload
49 views27 pages

A Study OF Artificial Immune Systems: Sunny Dubey Texas A&M University-Kingsville

This document summarizes a study on artificial immune systems. It provides an overview of biological immune systems, describing key features like affinity, diversity, and distributed operation. It also outlines three main algorithms inspired by the immune system: negative selection, clonal selection, and the immune network model. The document then discusses modeling immune system interactions and representations. It proposes a mathematical model for artificial immune systems and describes how it can be used for dynamic behavior arbitration, such as in robotics. Finally, it lists some potential applications of artificial immune systems like virus detection, intrusion detection, and pattern recognition.

Uploaded by

Sunny Dubey
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
49 views27 pages

A Study OF Artificial Immune Systems: Sunny Dubey Texas A&M University-Kingsville

This document summarizes a study on artificial immune systems. It provides an overview of biological immune systems, describing key features like affinity, diversity, and distributed operation. It also outlines three main algorithms inspired by the immune system: negative selection, clonal selection, and the immune network model. The document then discusses modeling immune system interactions and representations. It proposes a mathematical model for artificial immune systems and describes how it can be used for dynamic behavior arbitration, such as in robotics. Finally, it lists some potential applications of artificial immune systems like virus detection, intrusion detection, and pattern recognition.

Uploaded by

Sunny Dubey
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 27

A STUDY

OF
ARTIFICIAL IMMUNE SYSTEMS

Sunny Dubey
Texas A&M University-Kingsville

EEEN 5303 Independent Study

Spring 2011
OUTLINE
 Introduction

 Biological Overview

 Biologically inspired Immune Systems Algorithms

 Modeling the Biological Immune Systems

 String Matching Rules

 AIS Model

 Dynamic Behavior Arbitration using AIS

 Real World Applications


INTRODUCTION

 Complex and Adaptive system that differentiate between immune


cells as self-cells and non-self cells.

 Distributed and Parallel

 Two majors branches of the immune systems:


1) Innate Immune System
2) Adaptive Immune System

 Noticeable Information Processing capability


OVERVIEW OF BIOLOGICAL IMMUNE SYSTEMS

 The crucial features of the biological immune system are:


1) Affinity (matching)
2) Diversity
3) Distributed operation (no central mechanism)

 There are two immune cells that play important role in immune response:
1)B-cells (Bone Marrow),
2)T-cells (Thymus).

 Types of T-cells :
1) Helper T-cells
2) Killer T-cells
3) Suppressor T-cells
OVERVIEW OF BIOLOGICAL IMMUNE SYSTEMS (CONTD.)
 The immune system is divided into two heads:
1. Innate immunity (non-specific)
Congenital in nature
2. Adaptive immunity (specific )
It is crucial for learning and memory

 Such systems affect each other and linked to each other

 Again there are two types of adaptive immunity which are:


1. Hum-oral immunity
B-cells are responsible
2. Cell mediated immunity
T-cells are responsible
BIOLOGICALLY INSPIRED IMMUNE SYSTEM ALGORITHMS
(CONTD.)
There are three main immune system models:

1. Negative Selection Principle

Algorithm:

a. Self set P, Detector set M.

b. Generate set C.

c. Compare P and C.

d. Match found between P and C, discard and store other elements of P into
M.
BIOLOGICALLY INSPIRED IMMUNE SYSTEM
ALGORITHMS (CONTD.)
BIOLOGICALLY INSPIRED IMMUNE SYSTEM
ALGORITHMS (CONTD.)

2. Clonal Selection Algorithm (CLONALG)

The major two steps :

1. An antigen selects immune cells to proliferate.


Rate of proliferation  Affinity.

2. The mutation rate  1 / (affinity)


BIOLOGICALLY INSPIRED IMMUNE SYSTEM
ALGORITHMS (CONTD.)
BIOLOGICALLY INSPIRED IMMUNE SYSTEM
ALGORITHMS (CONTD.)

3. Immune Network Model

a) System is dynamic in absence of antigen – a proposal

b) Antibody recognizes other antibodies – the reason for dynamic behavior

c) Recognized molecular surfaces are coined as idiotopes.

d) Suppression of recognized antibody and stimulation of recognizing


antibody.
a.
BIOLOGICALLY INSPIRED IMMUNE SYSTEM
ALGORITHMS (CONTD.)
MODELING OF IMMUNE SYSTEMS
Shape-Space Model:

a) Describe the interactions between immune cell molecules and antigens


quantitatively

b) Recognition region around a antibody

c) Generalized shape of a molecule


MODELING OF IMMUNE SYSTEMS
MODELING OF IMMUNE SYSTEMS
Ag - Ab Representations and Affinities :
• Determine the distance measure that can be used to calculate the degree
of interaction between these molecules

• Mathematically, there are three ways to represent antibody-antigen pairs


and to determine their matching strength:

1. Euclidean shape-space
2. Manhattan shape-space
3. Hamming Shape-space
MODELING OF IMMUNE SYSTEMS

…Equation (1)

…Equation (2)

…Equation (3)
AIS MODEL
• The artificial immune system proposed by Farmer and Packard assumes
there is only one epitope for antibody and antigen.

• Antibody is represented as (p, e) where p represents paratope and e


represents epitope.

• Matching needs not to be exact. Matching threshold is defined as s ≤


min(le, lp), below which no antibodies react.

•
AIS MODEL (CONTD.)
On the basis of assumptions, the mathematical model for Artificial
Immune Systems is:
AIS MODEL (CONTD.)
Hypothesis for AIS Model:
AIS MODEL (CONTD.)
• In previous figure, (p1, e1) represents antibody ab1 that has recognized
by antibody ab2.

• So, ab’(n-1) is recognized by ab’n.


•
•Hypothesis: If ab’n is like original antigen, then it is like loop because
ab1 is going to recognize ab’n as shown in previous figure.

• This acts as memory for Artificial Immune Systems.


DYNAMIC BEHAVIOR ARBITRATION
• Three kinds of objects are assumed in proposed architecture by Akio
Ishiguro: 1. Predators. 2. Obstacles 3. Foods.
• Following assumptions are made:

1. Robot consumes energy Em.


2. If robot is captured, energy Ep is consumed.
3. If collision occurs, Eo energy is vanished.
4. Energy increases if robot gets food.
DYNAMIC BEHAVIOR ARBITRATION
Antibody Description:
DYNAMIC BEHAVIOR ARBITRATION
Prepared Antibody:
The antibody is activated if the robot detects the food in the front direction
and mid-range, and makes the robot move forward to pick it up.
DYNAMIC BEHAVIOR ARBITRATION
The dynamics for AIS Model proposed by Akio Ishiguro follows Farmer
and Packard mathematical formula.

ai = concentration of antibody,
mji = matching ratio between antibody I and j.
DYNAMIC BEHAVIOR ARBITRATION
DYNAMIC BEHAVIOR ARBITRATION
• Consider current energy level high, the antibodies 1, 2, 3, and 5 are
stimulated by the antigen.

• The concentrations of these very antibodies are incremented in


accordance to its antigen.

• The interaction within immune network’s antibodies is important. In the


end, antibody 5 is selected in previous figure.

• In the case of current energy level low, antibody 3 is selected.


APPLICATIONS OF AIS:
1. Virus Detection

2. Immunogenetic Approaches in Intrusion detection

3. Danger theory in Network Security

4. Robotics and Control

5. Pattern Recognition
QUESTIONS ?

You might also like