ECE 5730 Foundations of Neural Networks Summer I 2016 version 28 June 2016 The online version of this syllabus at has hyperlinks and will be updated as needed. Instructor Dr. Miller, Associate Professor of Electrical and Computer Engineering, Western Michigan University, College of Engineering and Applied Sciences, Parkview Campus, Room A-240, 269.276.3158, 269.276.3151 (fax),,. Office Hours Guaranteed office hours are posted on Dr. Miller’s door and. Please respect my office hours. Other times are available by appointment.

One to combine different logical systems in a principled way. Fibred neural. Keywords: Neural-Symbolic Integration. Fibring Systems, Recursion. Neural-Symbolic integration concerns the use of sym- bolic knowledge in the. AI (Cloete & Zurada 2000; d'Avila Garcez, Broda, &.

WMU Catalog Description ECE 5730 Foundations of Neural Networks, 3 hrs. Biological and artificial neural networks from an electrical and computer engineering perspective.

Neuron anatomy. Electrical signaling, learning, and memory in biological neural networks. Development of neural network circuit models. Artificial neural systems including multilayer feedforward neural networks, Hopfield networks, and associative memories. Electronic implementations and engineering applications of neural networks. Prerequisite Abilities You must be able to work independently on research projects and to write a professional quality written reports describing your project work.

Reading Assignments You must keep up with your reading. Exam/quiz questions may be developed from any assigned reading material. Much of the course material will require expanding your vocabulary; keep a list of new terms and their definitions.

As you read, think of questions to ask in class. ECE 5730 Course Learning Outcomes Graduates of ECE 5730 will exhibit (with most relevant ABET learning outcomes identified): 1. An understanding of the characteristics of intelligent systems (ABET: a,c); 2. An ability to develop numerical solutions of ordinary differential equations (ABET: a,e,k); 3. An understanding of basic neuron cell structure, anatomy, and functionality (ABET: a); 4. An understanding of neuron interactions via synaptic function (ABET: a); 5. An understanding of current knowledge of neural mechanisms that enable high level information processing in biological organisms (ABET: a); 6.

An ability to develop computer models of biological neuron(s) and biological neural networks (ABET: a,b,e,k); 7. An ability to design, analyze, and simulate circuits to model biological neuron(s) and biological neural networks (ABET: a,b,c,e,k); 8.

An understanding of common artificial neural network (ANN) architectures (ABET: a); 9. An understanding of adaptation and ‘learning’ in ANNs (ABET: a,e); 10. An understanding of classifier design, including the role of discriminant functions (ABET: a,e); 11. An ability to design and evaluate a multilayer feedforward neural network approximator or classifier (ABET: a,e); 12. A basic understanding of dynamical systems (ABET: a); 13.

An ability to perform a Lyapunov stability analysis (ABET: a,e); 14. An understanding of discrete and continuous feedback networks (ABET: a); 15. An understanding of associative memories (ABET: a); 16. An understanding of unsupervised learning techniques (ABET: a); 17. An ability to utilize computer simulations to study artificial neural networks (ABET: b,e,k); 18. An understanding of application areas for artificial neural networks, including pattern recognition, image processing, and signal processing (ABET: a,c,i); 19. Effective and ethical research methods with particular attention to proper citation techniques (ABET: f,k); and 20.

An ability to produce a concise summary of work performed using a standard journal paper format (ABET: k). Textbook and Materials Required: 1. Zurada, Artificial Neural Systems, PWS Publishing, Boston, 1992 (ISBN 0-314-93391-3). Available from the author, instructions for securing a copy to be provided in class. Otto Friesen and J. Friesen, NeuroDynamix II: Concepts of Neurophysiology Illustrated by Computer Simulations, Oxford University Press, 2010 (ISBN 978-0-19-537183-3). Scott Freeman, Biological Science, Prentice Hall, 2nd edition, 2005 (ISBN 0-13-140941-7): chapters 6 (“Lipids, Membranes, and the First Cells”), 45 (“Electrical Signals in Animals”), and 46 (“Animal Sensory Systems and Movement”) only.

Any version of this text is acceptable provided these chapters are present. Linear Technolog y, LTspice IV, available at no cost. You are responsible for ensuring access to a working copy. 5., MATLAB ® & SIMULINK ®. Download Clean Master Old Version For Android here. The student version is a tremendous value as this package includes many add-ons that must be purchased separately for use in a professional version.

References: (see Dr. Miller, might be put on reserve in ECE Department Office, check-out with WMU ID) 1.

Izhikevich, Dynamical Systems in Neuroscience: The Geometry of Excitability and Bursting, The MIT Press, Cambridge, Massachusetts, 2007. Simon Haykin, Neural Networks: A Comprehensive Foundation, IEEE Press, 1 st edition, 1994.

Smith, Microelectronic Circuits, Oxford University Press, 5 th edition, 1998. Maron, Numerical Analysis: A Practical Approach, Macmillan Publishing Co., Inc., 1982. Fuchs, From Neuron to Brain, Sinauer Associates, Inc., 2000. Scheinerman, Invitation to Dynamical Systems, Prentice Hall, 1996. Severance, System Modeling and Simulation, Wiley, 2001. Online References: 1. Teukolsky, W.

Vetterling, and B. Flannery, Numerical Recipes in C: The Art of Scientific Computing, Cambridge University Press, 2nd edition, 1992. Available online. Nave, HyperPhysics website,, outstanding physics tutorial/reference. Olivo, Biological Sciences 330/331 (Neurophysiology) website, Smith College,, See the links for videos shown in class.

Hodgkin and A. Driver For Labelflash Dvd Discs. Huxley, “A quantitative description of membrane current and its application to conduction and excitation in nerve,” J.

Physiol., no. 500-544, 1952. Squires, Instrumentation Electronics for an Integrated Electrophysiology Data Acquisition and Stimulation System, Master of Science in Electrical Engineering Thesis, Western Michigan University, December 2013, available at Course Policies Academic Honesty General: “You are responsible for making yourself aware of and understanding the University policies and procedures that pertain to Academic Honesty. These policies include cheating, fabrication, falsification and forgery, multiple submission, plagiarism, complicity and computer misuse. (The academic policies addressing Student Rights and Responsibilities can be found in the Undergraduate Catalog at and the Graduate Catalog at.) If there is reason to believe you have been involved in academic dishonesty, you will be referred to the Office of Student Conduct. You will be given the opportunity to review the charge(s) and if you believe you are not responsible, you will have the opportunity for a hearing. You should consult with your instructor if you are uncertain about an issue of academic honesty prior to the submission of an assignment or test.” — provided by the Professional Concerns Committee of the WMU Faculty Senate, bold face added, links updated.

Plagiarism: Plagiarism WILL NOT BE TOLERATED. See the Plagiarism Tutorial at to learn about plagiarism and how to properly cite sources.

• 96 Downloads • Abstract The effects of two insole materials within the shoe are compared using neural network analysis. Seven male subjects without locomotor disorders walk on a treadmill at a controlled speed and cadence wearing a common shoe and no socks, under three conditions; these are two types of insole of the same thickness, and a no insole condition. Pressure-related data from under the foot, within the shoe, are obtained by the MICRO-EMED system during walking. A back-propagation neural network is trained to associate sets of pressure-related data with the insole conditions. Subsequently neural network analysis is performed to reveal the abstract rules that govern the decision-making processes within the neural network, based on the synergistic interactions between the measured variables.

Data are also analysed using ANOVA. The neural network analysis finds trends in the way in which the trained neural network responds. The interpretation of those trends gives a delicate description of the dynamic behaviour of the insoles despite the fact that no significant differences are found using ANOVA. It is concluded that neural network analysis can distinguish between insole behaviour during use, even though these differences are not significantly different based on statistical tests.