1. Intelligent Systems Programming COMM2M Harry R. Erwin, PhD University of Sunderland

2. Purpose We will explore a few basic techniques of computational science. While computing is the study of algorithms, languages, and machines for solving problems, computational science focuses on a specific problem domain—in our case, intelligent behavior—and applies computing to scientific and engineering problems. The purpose of COMM2M (Intelligent Systems Programming) is to provide a foundation for your success in the Intelligent Systems MSc course. COMM2M (and COMM2E) are intended to help you understand how the topics presented in the other modules relate to science and computing.

3. Responsibility Dr. Harry Erwin is the Module Leader for COMM2M at the University of Sunderland. My primary research areas are computational neuroscience and auditory neuroethology—‘How bats do it.’ My PhD is in natural philosophy (experimental science), not computing. I also have 34 years of experience as a software systems engineer (mostly at TRW) on high-reliability systems. I supervise final year/MSc projects and PhD research in intelligent systems, security, and software engineering. E-mail: harry.erwin@sunderland.ac.uk or herwin@theworld.com. Neither are particularly reliable at present.harry.erwin@sunderland.ac.ukherwin@theworld.com We will use http://webct.sunderland.ac.uk as well.http://webct.sunderland.ac.uk

4. Goals Programming concepts for intelligent systems in using Java and Prolog. Basic concepts in the MatLab Neural Network and Image Processing Toolbox. The computational basis of the algorithms used in MatLab. Some GENESIS programming. Application of those programming tools to simple problems in intelligent systems.

5. Outline Twelve 2-hour lectures Twelve 90-minute tutorials One individual assignment. One group miniproject –You need an approved project proposal in two weeks.

6. My Informal Marking Criteria A first is supposed to mean that the student is clearly qualified to do a research degree with a significant computational component. An upper second is supposed to mean that the student is qualified to do a research degree with a significant computational component. A lower second is supposed to mean that the student may be qualified to do a research degree with a significant computational component. A third is supposed to mean that the student has reached a minimum level of skill in computational science. I expect almost all of you to earn a first or upper second.

7. Ethics and Academic Freedom Conrad Russell (Academic Freedom, Routledge, 1993) discusses the underlying principles: “the freedom for academics within the law to question and test received wisdom, and to put forward new ideas and controversial or unpopular opinions without placing themselves in jeopardy. It is the freedom to follow a line of research where it leads, regardless of the consequences, and the corresponding freedom to teach the truth as we see it, with suitable acknowledgement of views which differ from our own....”

8. Obligations of Scholars This discipline must be acquired by anyone who wishes to be a scholar: the willingness to speak the truth, not listening to pleas of convenience. This also implies duties of –truthfulness, –avoidance of willful error, and –avoidance of plagiarism. Remember the University is a community of scholars, not simply some buildings and facilities, and you are members of that community.

9. My Position on Collaboration The following are acceptable: –Cooperation in developing an understanding of project requirements. –Cooperation in the diagnosis of bugs and problems as long as the helper does not provide code solutions. –These are pedagogically valuable and allowed. The following are not acceptable: –Collaboration in developing project documentation and figures. Write your own report! –Collaboration in designing and coding your project. Come up with your own design and write your own code!

10. Break

11. Definitions Science is a dense network of careful experimental results, linked by theory. Engineering is the application of science to practical problems. Computational science is the application of computing to problems in science and engineering. Intelligent systems programming is computing in support of the engineering of lifelike systems.

12. The Role of Mathematics in Science Mathematics plays three roles in science: –A tool for handling numerical measurements –A tool for making numerical predictions –A notation for suggesting theory. Mainstream computing is weak in all three areas. –Numerical accuracy requires unexpected care. –Numerical predictions use techniques not taught in computing courses, and –Most languages are weak in supporting specialized notation The tools we will study are intended to overcome those problems.

13. Language Features and Science MatLab is designed to support easy engineering and scientific analysis, particularly when it involves matrices. Prolog eases the programming of rule-based systems. Java is deliberately designed for notational convenience. GENESIS is a programming language for neural systems.

14. Abstract Data Types Board presentation

15. Matrices Board Presentation

16. Questions?

17. Organize your Project Teams