Artificial Life - PowerPoint PPT Presentation


PPT – Artificial Life PowerPoint presentation | free to download - id: 600306-YmEzZ


The Adobe Flash plugin is needed to view this content

Get the plugin now

View by Category
About This Presentation

Artificial Life


RoadMap Define Artificial Life Artificial Immune Systems (AIS) ... Memories of the previous infections are retained Immune System Overview Immune Systems ... – PowerPoint PPT presentation

Number of Views:42
Avg rating:3.0/5.0
Slides: 30
Provided by: cseIitbA
Learn more at:


Write a Comment
User Comments (0)
Transcript and Presenter's Notes

Title: Artificial Life

Artificial Life
  • Perhaps the greatest significance of the
    computer lies in its impact on Mans view of
    himself The computer aids him to obey, for the
    first time, the ancient injunction Know thyself.
  • Herbert Simon (Nobel Laur., 1978)
    Group Members
  • Amit Arora
  • Manpreet Singh
  • Varun Garg
  • Vishaal Jatav

  • Define Artificial Life
  • Artificial Immune Systems (AIS)
  • Algorithms of Artificial Immune Systems
  • Applications of AIS
  • Danger Theory
  • Conclusion

What is Life ?
  • State of a functional activity and continual
    change, before death (defined complimentary as
  • Characterized by the capability to
  • Reproduce itself,
  • Adapt to an environment in a quest for survival,
  • Take Actions independent of exterior agents.

Definition of Artificial Life
  • Artificial Life is the study of man-made systems
    that exhibit behaviours characteristic of natural
    living systems (Langton)

  • Artificial life, also known as alife or a-life,
    is the study of life through the use of
    human-made analogues of living systems. ( )
  • A field of study that aims to discover the
    essential nature and universal features of
    "life" not only life as we currently know it,
    but life as it could be , whether on earth,
    within computers, or elsewhere , and in whatever
    shape or form that it may be found or made within
    our universe.

  • Agent Based Modelling
  • Economics , Ecological Resource Management
  • Infectious Diseases
  • Societal Structure and Dynamics
  • Consumer Market flows and Traffic Flows
  • Robotics
  • Gives rise to various classes of Algorithms
  • Swarm Intelligence
  • Artificial Immune Systems
  • Neural Networks
  • Genetic Algorithms

Ultimate Goal
  • Ultimate aim is to replicate all life processes
  • Commonly referred to as CYBORG
    Cybernetic Organism

Artificial Immune Systems
Artificial Immune Systems
  • Our bodys immune system is a perfect example of
    a learning system.
  • It is able to distinguish between good cells and
    potentially harmful ones(Antigens).

  • Artificial Immune Systems (AISs) are learning and
    problem solvers based on our own immune systems

  • AISs have been used to solve a wide variety of
    problems including
  • Computer Security,
  • Pattern Recognition,
  • Bridge Fault Detection
  • Data Mining

Immune System Overview
  • Pathogens are the germs that cause infection in
    the body
  • B cells are the detecting antibodies
  • T Cells are generated in B Cells that attack the
  • Memories of the previous infections are retained

Immune Systems - basic
  • Immune System principles
  • Immune network theory
  • Network of B cells
  • Negative Selection
  • Creation of detector set
  • Clonal Selection theory
  • Cloning of fit population

  • Parallel to a Self learning System
  • Initial set of cases used as training data
  • Self Learning System
  • Past experiences used

Artificial Immune Networks
  • Network of B cells
  • Artificial Recognition Ball (ARB) used to
    represent similar B cells
  • ARB Network creation procedure
  • Matching ARB found for the newly created B cell
  • B cell deleted if no matching ARB found
  • Empty ARBs deleted
  • Two B cells linked if the affinity value is less
    than Network Affinity Threshold (NAT)

Negative Selection
  • Concept of Self and non-Self
  • Antibodies should not react with body cells
  • Analogously detector set should not detect self
  • Procedure
  • Randomly generate detector Cells
  • Destroy any cell that matches self cells
  • Accept it otherwise

Clonal Selection
  • Fit cells are allowed to grow in number
  • Unfit are slowly removed
  • Cloning is directly proportional to the fitness
  • Mutation is inversely proportional to fitness
  • Procedure
  • calculate the fitness
  • select K best fit
  • clone them proportional to their fitness
  • mutate them inversely proportional to fitness

Fitness Calculation
  • Fitness Calculation

Overall Algorithm
  • Select Randomly a set P of cells
  • Apply negative selection
  • Removes self detecting cells
  • Clonal Selection
  • Calculate fitness
  • Select best fit ones
  • Clone and mutate accordingly
  • Network formation
  • Assign ARBs
  • Inter Link ARBs
  • Repeat till termination condition

Modelling a Problem in AIS
  • Symbols
  • Representation of measurements of the system
  • Pattern/Encoding
  • Simple structure of symbols.
  • Easy to sort.
  • Self Set
  • List of Patterns that represent normal
  • Obtained by training data.
  • Detector Set (B-Cell)
  • List of patterns that represent abnormal
  • Grows progressively by Learning.

AIS Application Bridge Fault Detection
  • Bridge Fault detection.
  • Bridge is analogous to the Human Body.
  • Vibrations caused in the bridge are antigens.
  • Self-Set contains safe patterns (e.g. cars,
    trucks etc).
  • Detector-Set (B-Cell) contains unsafe/dangerous
    vibrations (e.g. very heavy trailers, earthquake,

AIS Application in Data Mining
  • Movie Recommender System.
  • Server is the Human Body.
  • Incoming requests are the antigens.
  • Encoding is User (id1,score1,....,idn,scoren
  • Selection of Similarity Measure (correlation).
  • Create clusters based on the correlation measure.
  • AIS keeps on growing progressively by putting the
    new user into the relevant clusters.
  • Person interested in entertainment may also be
    interested in (say) cricket.

AIS Applications in Cyber Security
  • Intrusion Detection System
  • Computer Network is the Human Body.
  • Packets floating in the network are Antigens
  • Encode each data packet transferred as
  • ltprotocolgtltsource ipgtltsource portgt ltdest
    ipgtltdest portgt
  • Common antigens are known (common connections).
  • New connections activate the antigen, reporting
    the sysad.
  • If sysad smells a threat
  • He applies the patch.
  • Declares this antigen as SAFE to the Security
  • Network is Immunized.
  • Else (false alarm)
  • Declares this antigen as SAFE to the Security

(No Transcript)
Danger Theory
  • Self and Non-Self have been the classifications
    in the traditional theories of Immune System.
  • Danger Theory argues the concept as
  • IS differentiates only some self from some
  • The definition of self changes with changes in
    human body
  • Danger theory proposes
  • Foreign invaders that are dangerous stimulate
    danger signals by initiating cell stress or
  • IS manipulates these danger signals to recognise
    the danger zone and then evaluates the danger.

Concepts in Danger Theory
  • Danger Signals can be classified as
  • Apoptotic alerts(A alerts) are low-level alerts
    that could result from legit actions, but could
    also be the preparations of an attack.
  • Necrotic alerts(N alerts) relate to actual damage
    caused by a successful attack.
  • Danger Zone
  • Danger alerts are transmitted to the Immune
  • IS can quantify the degree of alert and indicate
    the strength of possible intrusion scenarios.
  • If there are strong indication of intrusion IS
    activates other sensors that are spatially,
    logically or temporally near the original sensor
    emitting the danger signal.
  • Danger Evaluation
  • IS makes a decision whether to activate a sensor
    or not.

Anomaly detection in file systems by DT
  • To the file systems, anomaly means unusual change
    or creation of the files, which can be caused by
    computer viruses, hacker intrusion, or some
    incidental errors.
  • Fails in traditional Anomaly detection
  • produces a large number of false flags (normal
    activities being tagged as abnormal).
  • unable to detect novel attacks.
  • Rapid development in computer hardware and
    software not taken into account.
  • Correlation in DT and File systems
  • Human Body The computer file system
  • A Signal Reading and writing of a file
  • N Signal Crashing of a file

Anomaly detection in file systems by DT (Contd...)
  • Correlation in DT and File systems
  • Danger Zone It receives the file information and
    creates a dynamic time neighbourhood of the
    changed files. It analyses the information in the
    neighbourhood, and sends the result to the danger
    evaluation module.
  • Danger Evaluation Gathering the information
    needed, a decision is made whether to send a
    danger signal or not. At the same time, a TCS
    (Threshold Control Signal) is sent back to the
    neighbourhood monitor, so that it can dynamically
    control its detection thresholds.

Anomaly detection in file systems by DT (Contd...)
  • Artificial Immune Systems
  • Highly distributed
  • Highly adaptive
  • Has the ability to continually learn about new
  • Artificial Life
  • Has Contributed a number of significant
  • Has applications in a number of fields.
  • Has innumerable unexplored areas

  • Alife wikipeida the free encyclopedia (
    http// )
  • Dasgupta, D (Editor). Artificial Immune Systems
    and Their Applications, ISBN 3-540-64390-7,
    Springer-Verlag, 1999.
  • Aickelin, U. and Cayzer, S., 2002c, The danger
    theory and its applicatioto artificial immune
    systems, in Proc. 1st Int. Conf. on Artificial
    Immune Systems (Canterbury, UK), pp. 141148.
  • Kim, J. and Bentley, P., 2002, Towards an
    artificial immune systems fornetwork intrusion
    detection an investigation of dynamic clonal
    selection,Proc. Congress on Evolutionary
    Computation 2002, pp. 10151020.
  • Dasgupta, D. Immunity-Based Intrusion Detection
    Systems A General Framework. In the Proceedings
    of the 22nd National Information Systems Security
    Conference (NISSC), October 18-21, 1999.

Thank You. Questions.