Textbook s

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

3. Introduction to Neural Networks and Fuzzy Logic Grade Policy • Final Grade = 20% Homework + 20% Quizzes + 30% Midterm Exam + 30% Final Exam + Extra Points • Homeworks will be given in fairly regular basis. The average of homework grades contributes 20% of final grade. • Homeworks are to be submitted on A4 papers, otherwise they will not be graded. • Homeworks must be submitted on time. If you submit late, < 10 min.  No penalty 10 – 60 min.  –20 points > 60 min.  –40 points • There will be 3 quizzes. Only the best 2 will be counted. The average of quiz grades contributes 20% of final grade.

4. Introduction to Neural Networks and Fuzzy Logic Grade Policy • Midterm and final exam schedule will be announced in time. • Make up of quizzes and exams will be held one week after the schedule of the respective quizzes and exams. • The score of a make up quiz or exam, upon discretion, can be multiplied by 0.9 (the maximum score for a make up is then 90). • Extra points will be given if you solve a problem in front of the class. You will earn 1, 2, or 3 points. • You are responsible to read and understand the lecture slides. I am responsible to answer your questions.

5. Neural Networks Introduction 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.

6. Neural Networks Introduction Experimental Modeling • Experimental modelingconsists of three steps: • The choice of model class • The choice of model structures (number of parameters, model order, time delay) • The calculation of the parameters and time delay. • The model may be chosen to be linear, nonlinear, or multi locally-linear. • A-priori (prior, previous) knowledge of the system to be modeled is required in most cases. • Artificial Neural Networks (or simply Neural Networks) offers a general solution for experimental modeling.

7. Neural Networks Introduction Experimental Modeling Using Neural Networks • 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. • It resembles the brain in two respects: • Knowledge is acquired by the network from its environment through a learning process. • Interneuron connection strengths, known as synaptic weights, are used to store the acquired knowledge.

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

9. Neural Networks Introduction Activation Function • Any continuous (differentiable) function can be used as an activation function in a neural network. • The nonlinear behavior of the neural networks is inherited from the used nonlinear activation functions. Tangent sigmoid function Logarithmic sigmoid function Radial basis function Linear function

10. Neural Networks Introduction Network Architectures Single layer feedforward network (Single layer perceptron) Input layer Output layer Multilayer feedforward network (Multilayer perceptron) Input layer Hidden layer Output layer

11. Neural Networks Introduction Network Architectures Diagonal recurrent networks Fully recurrent networks Input layer Hidden layer Output layer Input layer Hidden layer Output layer Delay element in arecurrent network

12. Neural Networks Introduction Network Architectures Elman’s recurrent networks Jordan’s recurrent networks

13. Neural Networks Introduction Preparation Assignment • Ensure yourself to install Matlab 7 in your computer, along with Matlab Simulink, Control System Toolbox, and Fuzzy Logic Toolbox. • Quizzes, Midterm Exam, and Final Exam will be computer-based.