In this paper, a method to teach advanced features of evolutionary algorithms (EAs), using a famous game known as Japanese puzzles is presented. The authors show that Japanese puzzles are constrained combinatorial optimization problems, that can be solved using EAs with different encodings, and are challenging problems for EAs. Other features, such as special operators and local search heuristics and its hybridization with genetic algorithms, can also be taught using these puzzles. The authors report an experience using this method in a course taught at the Universidad de Alcalaacute, Madrid, Spain

https://www.cs.bham.ac.uk/%7Exin/papers/T-Education2007.pdf

Teaching Advanced Features of Evolutionary Algorithms Using Japanese Puzzles

Algorithm animation has been applied to a huge variety of algorithms, and has been found to be of enormous benefit in assisting with their comprehension. However, the radical nature of several classes of algorithm has caused them to be largely overlooked in this regard; noticeable among these are genetic algorithms. In addressing this, we have used the XTANGO software system to develop a set of animation sequences designed to illustrate the behaviour of a genetic algorithm applied to a real-world problem. The resulting package is general enough to be adapted easily to a range of problems.

The use of animation to explain genetic algorithms

This paper presents the use of Java applets acting as a web-based interface to existing, platform dependent software tools. We present an example application called GAWebTutor which was constructed from a comprehensive genetic algorithm package and web-based Java components. In effect, we packaged an interactive genetic algorithms tutorial and made it available to anyone on the WWW. Furthermore, by building a GUI front end to an existing application in Java, we leveraged browser technology to overcome platform compatibility problems. A blending of Java and platform dependent code can successfully create user friendly, portable, more versatile applications that take full advantage of all available software components. Using our approach, educators can make their educational programs or other software tools available to anyone through the WWW.

Using Java to develop Web based tutorials

ion, which is based on the experimental scientific method, also consists of four steps used for investigating a phenomenon: 1. form a hypothesis; 2. construct a model and make a prediction; 3. design an experiment and collect data; and 4. analyse results. Design, which is based on engineering, consists of four steps that are used for constructing a system to solve a given problem: 1. state requirements; 2. state specifications; 3. design and implement the system; and 4. test the system. Computer science focuses on theory and abstraction, whereas computer engineering focuses on abstraction and design, as Denning et al. (1989) have stated. From now on, computer science and computer engineering will be combined under the name of computer science. Different methods play an important role in computer science instruction, since students are taught to recognise problems, find solutions to these problems, and finally, implement these solutions in a computer format. Good knowledge of both mathematics and logic are necessary for computer scientists since implementing complex computer systems requires ability both to understand abstract problems and to draw conclusions. To quote Denning et al. (1989): The discipline of computing is the systematic study of algorithmic processes that describe and transform information: their theory, analysis, design, efficiency, implementation, and application. The fundamental question underlying all of computing is, “What can be (efficiently) automated?” At the university level, computer science education emphasises general abilities to solve computer tasks and good general education on available computer tools and techniques, rather than practical skills. Therefore, instruction of computer science mostly consists of teaching abstract concepts and methods, though practical skills cannot be totally forgotten. Computer science is a fast developing discipline, which causes great challenges to educators. Changes both in hardware and software force frequent assesment of the computer science curriculum. This means that the curriculum should be updated to cover new concepts and methods. These changes reflect also to CAI courses on computer science that should be flexible to update. Since changes within the discipline happen fast, updating is necessary in order not to waste lots of design and implementation effort put into CAI courses on this domain. 4.2. Criteria based on subject 35 The domain of computer science contains a wide range of diverse topics. It can be divided roughly into the following subdomains (Denning et al., 1989): • algorithms and data structures • programming languages • architecture • numerical and symbolic computation • operating systems • software methodology and engineering • databases and information retrieval • artificial intelligence and robotics • human-computer communication This is not the only possible division but it describes the domain very well and is used for different purposes, such as for classification of literature in Computing Reviews or for designing computer science curricula. Each of these subdomains has its elements of theory, abstraction and design that define the main features of it. The weights of theory, abstraction and design differ within each subdomain, that is, some subdomains are considered more practical than theoretical. As an example, algorithms and data structures, numerical and symbolic computation, and software methodology and engineering have a theoretical nature, whereas programming languages, architecture, and operating systems are aimed more at practice. The next discussion of some of these subtopics is based on a report of Denning et al. (1989). Algorithms and data structures have a very strong theoretical background, which consists of computability theory, computational complexity theory, time and space bounds, levels of intractability, parallel computation, probabilistic algorithms, cryptography, and some supporting areas (e.g., graph theory, combinatorics, calculus, and semantics). Computer science students should be familiar with this theoretical background before they can fully understand how and why algorithms are designed the way they are, and how algorithms work. In addition, the theory of algorithms and data structures gives students fundamental concepts and guidelines to abstract and design their own algorithms. Algorithms and data structures form the basis for all computing, thus, students necessarily need to know them. The theoretical basis of programming languages consists of formal languages and automata, Turing machines, Post systems and λ-calculus, formal semantics, and supporting areas, such as, predicate logic and temporal logic. This theoretical part often has a minor role in a computer science curriculum that prefers to concentrate on introducing different programming languages. This kind of attitude is harmful since without proper theoretical knowledge students are unable to understand the fundamental questions of programming language area, namely, what possible organisations presented by the language are, how these abstractions are implemented on computers, and what notation can be used effectively 36 4. Criteria for evaluation of CAI courses on computer science and efficiently to specify what the computer should do. Programming languages are useful for giving substance to algorithms if the theoretical background is not forgotten. The area of architecture aims at organising hardware into efficient, reliable systems, such as hardware units for fast computation (e.g., arithmetic function units), efficient methods of both storing and recording information and detecting and correcting errors, computer aided design (CAD) systems, and supercomputers (e.g., Cray and Cyber machines). Theory behind these aims is based on Boolean algebra, switching theory, coding theory, finite state machine theory, and supporting areas, such as queuing, reliability theory, discrete mathematics, and number theory. Architecture is easily misunderstood merely as pure hardware. This misconception should be corrected by emphasising the fundamentals of computer architecture in instruction. Students often complain about the amount of mathematics included in computer science curriculum. It should be explained why they need such skills and knowledge. Numerical and symbolic computation focus on solving several problems, such as how we can accurately approximate continuous or infinite processes by finite discrete processes, how rapidly a given class of equations can be solved for a given level of accuracy, and how symbolic manipulations on equations (e.g., integration, differentiation, and reduction to minimal terms) can be carried out. The theories needed for understanding the meaning of and reaching these aims are number theory, linear algebra, numerical analysis, nonlinear dynamics, and as supporting areas calculus, real analysis, complex analysis, and algebra. Students need to learn at least the basics of these theories in order to be able to understand, design and implement algorithms solving mathematical problems. To quote Repo (1996), “a deep conceptual understanding is important in the process of learning algorithmic skills.” Database and information retrieval systems emphasise several topics, such as choosing appropriate modeling concepts to represent data elements and their relationships, combining basic operations into effective transactions, translating high-level queries into highperformance programs, protecting data against unauthorised access, disclosure, or destruction, and indexing and classifying text for efficient retrieval. The theoretical basis of this subdomain is also very extensive including relational algebra and relational calculus, dependency theory, concurrence theory, statistical inference, sorting and searching, performance analysis, and, as supporting theory, cryptography. Database and information retrieval systems are needed more and more in everyday life. Therefore, computer science students have to know this area well if they want to design and implement even better systems in future. Computer science is a versatile domain that cannot be taught using solely one method. Some subdomains are clearly theoretical, others practical or mathematical, though all these aspects can be found in every subdomain. Therefore, a mixture of these different aspects needs to be considered when designing computer science curricula. 4.3. Instructional criteria 37 4.3 Instructional criteria Computer-aided instruction should be student-centered and based on student’s needs. Consequently, CAI courses should be interactive and controlled together by a student and computer. These demands, when they are fulfilled, promote student’s motivation, whose importance as a part of successful instruction is strongly emphasised in many human learning theories. Another thing to remember is that instruction should be based on students’ previous knowledge. This is important because then students are able to connect new information to their existing knowledge which strengthens their learning. At first it should be confirmed the general view and basic concepts of the instructed domain are familiar to students. Together these two facilitate students’ learning by offering them an appropriate knowledge base on which they are able to combine novel information. When students have a good general view of the domain they have better possibilities to recognise the connections between the domain and the novel information. As a result, every CAI course should clearly announce the prerequisites needed to understand the contents of the course. As an example, an introductory CAI course on the domain could give the required general view on the domain and its concepts, and no prerequisites on this domain should be necessary for understanding this course. Other CAI courses on subtopics of that domain could announce that the introductory course, or knowledge equivalent to 

Cai in cs

An Indexed Bibliography of Genetic Algorithms Basics, Reviews, and Tutorials

In this paper we discuss the design and implementation of GATutor, a graphical tutorial system for genetic algorithms (GA). The X Window/Motif system provides powerful tools for the development of a user interfaces with a familiar feel and look. We implemented the Traveling Salesman Problem (TSP) and the Set Covering Problem (SCP) as two example GA problems in the tutorial. The TSP problem uses an order-based chromosome representation (permutation of n objects), while the SCP uses bit strings. The user has numerous buttons to select the GA parameters. These include (a) type of initial population: random or from a file, (b) mode: steady-state or generational, (c) population size, (d) maximum number of generations or trials, (e) generation gap, (f) selection mode, (g) selection bias, (h) selection of the crossover operation from a choice of several possibilities, (i) mutation method, (j) mutation rate, (k) replacement method, (l), elitism, etc. The user has the ability to do astep by step execution or to do a continuous run. The screen layout provides visual representation of the chromosomes in the population with the ability to scroll. This gives the user the option of varying one or two GA parameters to visually see the effect on the algorithm. One of most important features of this tutorial is the set of help screens that explain, with examples, all of the options for each of the GA parameters. This package has already been very useful for teaching the fundamental features of GAs in many different courses, and it has been very valuable in our GA research projects.

Charles Prince

Papers

GATutor: a graphical tutorial system for genetic algorithms