Introduction to Neural Networks and Fuzzy Logic Lecture

  • Slides: 15
Download presentation
Introduction to Neural Networks and Fuzzy Logic Lecture 1 Dr. -Ing. Erwin Sitompul President

Introduction to Neural Networks and Fuzzy Logic Lecture 1 Dr. -Ing. Erwin Sitompul President University http: //zitompul. wordpress. com 2 0 1 8 President University Erwin Sitompul NNFL 1/1

Introduction to Neural Networks and Fuzzy Logic Textbooks Textbook: “Neural Networks. A Comprehensive Foundation”,

Introduction to Neural Networks and Fuzzy Logic Textbooks Textbook: “Neural Networks. A Comprehensive Foundation”, 2 nd Edition, Simon Haykin, Prentice Hall, 1999. “Fuzzy Systems Theory and Its Application”, Toshiro Terano et. al. , Academic Press, 1992. President University Erwin Sitompul NNFL 1/2

Introduction to Neural Networks and Fuzzy Logic Grade Policy Final Grade = Academic +

Introduction to Neural Networks and Fuzzy Logic Grade Policy Final Grade = Academic + 9 Values + Extra Points Academic = 6% Notes + 16% Homework + 16% Quizzes + 26% Midterm Exam + 26% Final Exam 9 Values = 6% Peer Assessment + 4% Lecturer Assessment n Your handwritten note will be collected after Quiz 3, and given back to you on the next class. n Handwritten note contributes 6% of final grade. n Homeworks will be given in fairly regular basis. The average of homework grades contributes 16% of final grade. n Written homeworks are to be submitted on A 4 papers, otherwise they will not be graded. n Homeworks must be submitted on time, on the day of the next lecture, 10 minutes after the class starts. Late submission will be penalized by point deduction of – 10·n, where n is the total number of lateness made. President University Erwin Sitompul NNFL 1/3

Introduction to Neural Networks and Fuzzy Logic Grade Policy n There will be 3

Introduction to Neural Networks and Fuzzy Logic Grade Policy n There will be 3 quizzes. Only the best 2 will be counted. The average of quiz grades contributes 16% of final grade. n Make up of quizzes must be held within one week after the schedule of the respective quiz. n Midterm exam (26%) and final exam (26%) follow the schedule released by AAB (Academic Administration Bureau). Date of the lecture when the homework is issued • Heading of Written Homework Papers (Required) President University Erwin Sitompul NNFL 1/4

Introduction to Neural Networks and Fuzzy Logic Grade Policy n Extra points will be

Introduction to Neural Networks and Fuzzy Logic Grade Policy n Extra points will be given if you solve a problem in front of the class. You will earn 1 or 2. n Lecture slides can be copied during class session. It is also available on internet. Please check the course homepage regularly. http: //zitompul. wordpress. com n The use of internet for any purpose during class sessions is strictly forbidden. President University Erwin Sitompul NNFL 1/5

Neural Networks Introduction to Neural Networks Validation: • Generally, means confirming that a product

Neural Networks Introduction to Neural Networks Validation: • Generally, means confirming that a product or service meets the needs of its users. • Testing whether the mathematical model is good enough or not to describe the empirical phenomenon. President University Erwin Sitompul NNFL 1/6

Neural Networks Introduction Experimental Modeling n Experimental modeling consists of three steps: 1. The

Neural Networks Introduction Experimental Modeling n Experimental modeling consists of three steps: 1. The choice of model class 2. The choice of model structures (number of parameters, model order, time delay) 3. The calculation of the parameters and time delay. n The model may be chosen to be linear, nonlinear, or multi locallylinear. n A-priori (prior, previous) knowledge of the system to be modeled is required in most cases. n Artificial Neural Networks (or simply Neural Networks) offers a general solution for experimental modeling. President University Erwin Sitompul NNFL 1/7

Neural Networks Introduction Experimental Modeling Using Neural Networks n A neural network is a

Neural Networks Introduction Experimental Modeling Using Neural Networks n A neural network is a massively-parallel distributed processor made up of simple processing unit, which has natural propensity for storing experiential knowledge and making it available for use. n It resembles the brain in two respects: 1. Knowledge is acquired by the network from its environment through a learning process. 2. Interneuron connection strengths, known as synaptic weights, are used to store the acquired knowledge. President University Erwin Sitompul NNFL 1/8

Neural Networks Introduction Biological and Artificial Neuron Structure of Biological neuron Activation function Structure

Neural Networks Introduction Biological and Artificial Neuron Structure of Biological neuron Activation function Structure of Artificial neuron President University Erwin Sitompul NNFL 1/9

Neural Networks Introduction Activation Function n Any continuous (differentiable) function can be used as

Neural Networks Introduction Activation Function n Any continuous (differentiable) function can be used as an activation function in a neural network. n The nonlinear behavior of the neural networks is inherited from the used nonlinear activation functions. Linear function President University Tangent sigmoid function Logarithmic sigmoid function Erwin Sitompul Radial basis function NNFL 1/10

Neural Networks Introduction Network Architectures Single layer feedforward network (Single layer perceptron) Input layer

Neural Networks Introduction Network Architectures Single layer feedforward network (Single layer perceptron) Input layer Output layer Multilayer feedforward network (Multilayer perceptron) Input layer President University Erwin Sitompul Hidden layer Output layer NNFL 1/11

Neural Networks Introduction Network Architectures Diagonal recurrent networks Input layer Hidden layer Output layer

Neural Networks Introduction Network Architectures Diagonal recurrent networks Input layer Hidden layer Output layer Fully recurrent networks Input layer Hidden layer Output layer Delay element in a recurrent network President University Erwin Sitompul NNFL 1/12

Neural Networks Introduction Network Architectures Elman’s recurrent networks President University Jordan’s recurrent networks Erwin

Neural Networks Introduction Network Architectures Elman’s recurrent networks President University Jordan’s recurrent networks Erwin Sitompul NNFL 1/13

Neural Networks Introduction Preparation Assignment n Ensure yourself to install Matlab 7 or newer

Neural Networks Introduction Preparation Assignment n Ensure yourself to install Matlab 7 or newer in your computer, along with Matlab Simulink, Control System Toolbox, and Fuzzy Logic Toolbox. n Quizzes, Midterm Exam, and Final Exam will be computer-based. President University Erwin Sitompul NNFL 1/14

Neural Networks Introduction Homework 1 A n This is an individual work. n Conduct

Neural Networks Introduction Homework 1 A n This is an individual work. n Conduct a literature research and prepare a short Power. Point presentation about the applications and implementations of neural networks in area of: 1. Manufacturing (Tamara) 2. Robotics (Andre) 3. Medics (Keanu) 4. Military (Fikri) 5. Administration/ Data Mining (Rudy) 6. Sports/ Entertainment (Fadhilla) 7. Recognition/ Identification (Alief) 8. E-Commerce (Maulidya) n Show the structure of the neural networks used in your presentation n You will be given 15 minutes time for presentation on Monday, 22 January 2018. President University Erwin Sitompul NNFL 1/15