Scribd Logo
Upload

Welcome to Scribd!

  • 390 views

    Unit 1 - Introduction

    Uploaded by

    Pratiksha Kamble
    This document provides an overview of the course "Soft Computing" taught by Jagadish Kashinath Kamble. The key points covered include: - An introduction to soft computing techniques like neural networks, fuzzy logic, and genetic algorithms. - Six units that will be covered related to these topics as well as applications and advances. - Course objectives like generating ability to build neural networks and apply evolutionary algorithms. - A brief history of soft computing including its definition and goals of approximating real-world problems.

    Copyright:

    © All Rights Reserved
    Download as PPT, PDF, TXT or read online from Scribd

    Unit 1 - Introduction

    Uploaded by

    Pratiksha Kamble
    390 views25 pages
    This document provides an overview of the course "Soft Computing" taught by Jagadish Kashinath Kamble. The key points covered include: - An introduction to soft computing techniques like neural networks, fuzzy logic, and genetic algorithms. - Six units that will be covered related to these topics as well as applications and advances. - Course objectives like generating ability to build neural networks and apply evolutionary algorithms. - A brief history of soft computing including its definition and goals of approximating real-world problems.

    Original Title

    Unit 1- Introduction.ppt

    Copyright

    © © All Rights Reserved

    Available Formats

    PPT, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document provides an overview of the course "Soft Computing" taught by Jagadish Kashinath Kamble. The key points covered include: - An introduction to soft computing techniques like neural networks, fuzzy logic, and genetic algorithms. - Six units that will be covered related to these topics as well as applications and advances. - Course objectives like generating ability to build neural networks and apply evolutionary algorithms. - A brief history of soft computing including its definition and goals of approximating real-world problems.

    Copyright:

    © All Rights Reserved
    Download as PPT, PDF, TXT or read online from Scribd
    Download as ppt, pdf, or txt
    390 views25 pages

    Unit 1 - Introduction

    Uploaded by

    Pratiksha Kamble
    This document provides an overview of the course "Soft Computing" taught by Jagadish Kashinath Kamble. The key points covered include: - An introduction to soft computing techniques like neural networks, fuzzy logic, and genetic algorithms. - Six units that will be covered related to these topics as well as applications and advances. - Course objectives like generating ability to build neural networks and apply evolutionary algorithms. - A brief history of soft computing including its definition and goals of approximating real-world problems.

    Copyright:

    © All Rights Reserved
    Download as PPT, PDF, TXT or read online from Scribd
    Download as ppt, pdf, or txt
    of 25

    Soft Computing

    Jagadish Kashinath Kamble (M.Tech, IIT Kharagpur)


    Dept. of Information Technology,
    Pune Institute of Computer Technology, Pune

    1
    About the Course

    • Course Instructor: Jagadish Kashinath Kamble


    • A3-310, [email protected]
    •Online Lectures / Classroom Lectures
    •Evaluation in the theory course (100 Marks):

    In-Semester-30%

    End-semester -70%

    • Attendance REALLY matters


    •Important for understanding the course
    •Classes will involve both Slides + Board (to roughly equal degrees)

    • Slides will be made available at PICT Website and PIAZZA Class Group.
    Books
    Books (Text Books and Reference Books)
    1.S. N. Sivanandam, S. N. Deepa, Principles of Soft Computing, Wiley publications, 2nd
    Edition.
    2.J. S. R. Jang, C. T. Sun, E. Mizutani, 'Neuro-Fuzzy and Soft Computing- A computational
    approach to Learning and Machine Intelligence' PHI
    3.David E. Goldberg, Genetic Algorithms, Pearson Education, 2nd Edition, ISBN:
    9788120322431, ISBN: 9780201157673.
    4. Satish Kumar, Neural Networks - A Classroom Approach, Tata McGraw Hill,2nd Edition,
    ISBN:1259006166.
    5. Timothy J. Ross, Fuzzy Logic with Engineering Applications, Wiley India, 3rd Edition, ISBN:
    9788126531264.
    6. Samir Roy, Udit Chakroborthy, Introduction to soft computing neuro-fuzzy and genetic
    algorithm ,Person Education, 1st Edition.
    Pre-requisites
    To extract the maximum from the course, the following prerequisites are must.
    •Linear Algebra and Calculus
    •Probability Theory
    •A strong mathematical background
    •Proficiency with algorithms
    •Programming skills in C, C++, or Java, MATLAB, etc
    •Critical thinking and problem solving skills
    Course Objectives

    The learning/course objectives of this course are to


    1.Identifying Soft computing techniques and their roles in problem solving.
    2.Generate an ability to build neural networks for solving real life problems.
    3.Conceptualize fuzzy logic and its implementation for various real world applications.
    4.Apply evolutionary algorithms and Fuzzy logic to solve the problems.
    5.Design soft computing systems by hybridizing various other techniques.
    Learning Outcomes

    After completing this course, students will be able to:


    1.Tackle problems of interdisciplinary nature.
    2.Find an alternate solution, which may offer more adaptability, resilience and
    optimization.
    3.Gain knowledge of soft computing domain which opens up a whole new career
    option.
    4.Tackle real world research problems.
    Course Contents

    Unit I. Introduction
    Unit II. Neural Network Overview
    Unit III. Neural Network Architecture
    Unit IV. Fuzzy logic and Fuzzy Systems
    Unit V. Genetic Algorithms
    Unit VI. Advances in Soft Computing
    Unit -I : Introduction
    • Basic concepts of Soft Computing
    • Historical Developments , Definitions
    • Soft Computing Characteristics
    • Problem Solving– Strengths and Weaknesses
    • Constitutes of Soft Computing :
    – Neural Computing
    – Fuzzy Logic and Computing
    – Evolutionary Computing
    – and Genetic Algorithms
    – Probabilistic Reasoning
    Unit -II : Neural Network Overview
    • Fundamentals:
    – Biological Neurons
    – Model of Artificial Neuron.
    • Neural Network Architectures:
    – Single Layer Network
    – Multi Layer Feed Forward Neural Networks
    – Feedback Networks.
    • Perceptron Model and Learning in Perceptron
    – Limitation of Learning in Perceptron
    • Error Back Propagation learning in Multilayer FFNN (EBF- MLFFNN)
    – Performance Issues of EBP algorithm for MLFFNN.
    Unit -III : Neural Network Architecture
    • Complex Architectures Learning:
    • Competitive Learning-
    – Self Organizing Maps
    • Hebbian Learning-
    – Hopfield Networks
    • Boltzmann Machines
    • Adaptive Resonance Theory (ART) Networks
    • Bayesian Neural Networks
    • Deep Learning Architecture of Neural Networks
    • Applications of Neural Networks
    Unit -IV : Fuzzy Logic and Fuzzy system
    • Fuzzy Logic
    • Fuzzy Sets and Operations
    • Fuzzy Relations
    • Fuzzy Arithmetic and Fuzzy Measures
    – Fuzzy to Crisp Conversions:
    – Lambda Cuts for fuzzy sets,
    – Fuzzy Relations,
    – Defuzzification Methods.
    – Fuzzy Rules and Reasoning,
    • Fuzzy Inference Systems,
    • Mamdani Fuzzy Models – Sugeno Fuzzy Models
    • Applications of Fuzzy Modeling for Decision Making
    Unit -V : Genetic Algorithm
    • Introduction Encoding
    • Operators of Genetic Algorithm
    • Basic Genetic Algorithm
    • Simple GA
    • Crossover and Mutation
    • Multi-objective Genetic Algorithm (MOGA).
    • Genetic algorithms in search and optimization
    • Ant colony optimization (ACO)
    • Particle Swarm Optimization (PSO).
    • Applications of GA for Clustering
    Unit -VI : Advances in Soft Computing
    • Soft Computing Paradigms
    – Hybrid Approaches
    – Neuro-Fuzzy modeling
    • Genetic Algorithm Based Backpropagation Network,
    • Fuzzy logic based Back propagation,
    • Fuzzy Logic Controlled Genetic Algorithms, Simplified Fuzzy ARTMAP.
    Soft Computing

    Unit I
    Introduction

    Jagadish Kashinath Kamble (M.Tech, IIT Kharagpur)


    Dept. of Information Technology,
    Pune Institute of Computer Technology, Pune
    • ‘We are drowning in information and starving for knowledge’.(Ruthorford,
    1985).
    • Real life Problems

    • Computers and Digital Evolution

    • Traditional Mathematical Model

    • Data Mining

    • Intelligent Data Analysis


    – Mimicking the natural process of evolution

    • Bioinformatics
    • Learning from the data

    • Soft Computing

    • Expert Systems

    • Natural Computations
    – Cellular automata inspiration from self reproduction

    – Evolutionary computational methods due to Darwinian evolution of species.

    – Neural Computations mimicking the function of brain

    – Swarm intelligence methods motivated by the group behavior of organisms

    • Nature
    Soft Computing

    • Soft Computing (SC) is an emerging area in Computer Science that is tolerant to


    imprecise and uncertain problems with partial truth to achieve an approximate,
    robust and low cost optimal solution.
    • stock market predictions in business, computer aided diagnosis in medical,
    handwriting recognition in fraud detection , image retrieval, biometric application in
    image processing etc. and the list go on.
    • Initiated by L.A. Zadeh in 1981 “What is soft Computing”

    • Soft Computing into one multidisciplinary system as the fusion of the fields of
    Fuzzy Logic, Neuro-Computing, Evolutionary and Genetic Computing, and
    Probabilistic Computing
    Soft Computing

    • Aim of Soft Computing

    • Definitions :
    – Soft Computing is an emerging approach to computing which parallel
    the remarkable ability of the human mind to reason and learn in a environment of uncertainty and
    imprecision
    Computing

    • Hard Computing

    • Soft Computing
    – Fuzzy set / Fuzzy Logic

    – Neural Networks /

    – Genetic Algorithms /Evolutionary Computation

    • Soft Computing vs Hard Computing


    – Biological Process

    – Heuristics

    –
    Example: Hard Computing vs Soft Computing
    • Problem: If we need to find out whether X is honest. A hard computing algorithm
    would give an answer that is YES or No. (1 or 0 in binary) whereas a SC technique
    (Fuzzy Logic) would give an answer with membership degree such as extremely
    honest (1), very honest (0.85), sometimes honest (0.35), extremely dishonest (0.00),
    like a human

    • Prediction Problem: Will tomorrow rain? Hard Computing would solve the problem
    by using statistical techniques which basically rely on user-driven hypothesis test
    that user defines variables, functions and type of interaction. These type of
    information can be manipulated and influence models. Whereas the hypothesis of
    Soft Computing techniques would scan all the predictor variables automatically and
    identify some patterns in the data that help to come up with accurate predictions.
    Such models have no chance to miss a notice of unexpected and potentially
    important variables, thereby lead to better accuracy.
    Goals of Soft Computing

    • Computational Intelligence

    • Develop intelligent machines to provide solutions to real world problems

    • tolerance for Approximation, Uncertainty, Imprecision, and


    Partial Truth
    – Approximation : here the model features are similar to the real ones, but
    not the same.
    – Uncertainty : here we are not sure that the features of the model are the
    same as that of the entity (belief).
    – Imprecision : here the model features (quantities) are not the same as that
    of the real ones, but close to them
    Historical Development

    • Computers used Everywhere

    • Real life problems

    • Zadeh, father of Fuzzy Logic

    • 1955, John McCarthy organized Summer Research Project on AI

    • In the mid of 1980s Zadeh focused on building systems or making


    computers think like humans.
    • Soft Computing

    • 1981, Paper “Soft Data Analysis”, Soft Computing


    Historical Development

    • Soft Computing (Zadeh in 1981)


    – Evolutionary Computing (Rachenberg 1960)

    – Neural Network (McCulloch, 1943)

    – Fuzzy Logic (Zadeh, 1965)

    • Evolutionary Computing
    – Genetic Programing (Koza, 1992)

    – Evolution Stratergies (Rechenberg, 1965)

    – Evolutionary Programming(Fogel, 1962)

    – Genetic Algorithms( Holland, 1970)


    The Soft Computing – development
    history

    SC = EC + NN + FL
    Soft Evolutionary Neural Fuzzy
    Computing Computing Network Logic
    Zadeh Rechenberg McCulloch Zadeh
    1981 1960 1943 1965
    EC = GP + ES + EP + GA
    Evolutionary Genetic Evolution Evolutionary Genetic
    Computing Programming Strategies Programming
    Algorithms
    Rechenberg Koza Rechenberg Fogel Holland
    1960 1992 1965 1962 1970

    You might also like