Reference no: EM132333597

Assignment - Programming Report

Overview of Assessment - This assignment assesses the following module learning outcomes:

1. To apply an appropriate technique(s) to a given problem.

2. Formulate a problem such that it is amenable to modern Artificial Intelligence techniques.

3. Appraise the usefulness of various techniques for particular situations.

Broadly speaking, the assignment requires you to write a report on your attempts to solve a set of optimization problems as effectively as possible using any form of evolutionary intelligence covered on the course. This requires you to write your own code, in a language of your choice, building upon your own genetic algorithm code written and developed in the first few lab sessions.

This is an individual assignment. Working on this assignment will help you to develop your understanding of how learning can be seen as a search process and how the parameters controlling search techniques affect their ability to solve tasks. If you have questions about this assignment, please email the module leader.

Task Specification -

Three functions are provided, the first two of which should be tackled using a binary encoding of the variables and the third using a real valued encoding. The task is to evolve a system that finds the maximum (or minimum) of each of the functions. To pass the assignment, you must implement a system that successfully searches a binary encoded problem space and demonstrate the effects of parameter changes, preferably through graphs and including your understanding of what is happening.

Include a research section which briefly reviews optimization and the use of evolutionary computing techniques for such problems in particular. In an experimentation section describe the encoding used for the optimization, show example runs and solutions found.

Function 1: f(x) = x² Where x is an integer in the range 0-255, i.e., 0 ≤ x ≤ 255

Function 2: f(x, y) = 0.26.( x² + y²) - 0.48.x.y Where -15 ≤ x, y ≤ 15

Function 3: f(x) = 10n + _i=1∑ⁿ x_i² - 10.cos(2π.xi) Where -5.12 ≤ x_i ≤ 5.12, and use n=10, 20

The worked example of solving f(x) = x² in the first lecture on evolutionary algorithms will provide you with a starting point for Function 1. Function 1 is a maximization problem, whereas Functions 2 and 3 are minimization problems. More marks will be given to the effective use of more sophisticated approaches, particularly for the real-valued function. Alternatively, once you have solved the problems with your own code, you may use freely available software to compare performance with other optimization approaches - Tabu search, simulated annealing, etc.

Deliverables - Depending on font size, and line spacing, around 5 pages is a reasonable target length. The intention is your hand-in approximates to a research paper - please use the template provided. Please include commented source code as a printed Appendix or via an active link.