程序代写代做代考 algorithm Introduction

Introduction

School of Information Science & Engineering

7168: Soft Computing
7197: Soft Computing PG

Soft Computing / PG: GA assignment

Genetic Algorithms: Assignment 2016
Aid Convoy Trucks
45 Marks
You should demonstrate your answer in the tutorial in week 13, either the 31st October or 3rd November. The written report and code submission due on Saturday 5th November just before midnight via moodle.

Genetic Algorithms is one of a set of meta-heuristics that have been applied to optimization problems that do not have algorithmic solutions. These problems have to rely on heuristics that are tailored to the problem, or meta-heuristics – generalised heuristics that can be applied to a variety of problems.

Your assignment is to investigate how well GA performs on one particular problem, the packing of 3 aid trucks to cyclone ravaged Darwin.
In the classes supplied there is a list of items that will be loaded onto the 3 trucks.
You need to code the following (This is a minimum – feel free to rewrite my shell)

1. Your gene structure/class

2. Your genome structure/class

3. Your mutation strategy/system/method(s)

4. Finish the method showBest() to display the answers
5. Anything else needed to make the program run as a GA; (eg initialise the population with random genes)

Your need to demonstrate your program in week 13 at your tutorial, either the 31st October or 3rd November.
You will also need to submit a short report which should include the following:
· Your gene structure

· Your genome structure

· Your mutation strategy

· Thing that you found interesting or difficult during development of your GA
· The effect of varying population size, number of generations, mutation rate
There is no fixed size for your report, but I would expect it to be between 1 – 3 pages.

You will also need to zip the code for your GA project and submit it to moodle.

Marking Criteria:

To get a Pass (23+ marks)
The Program must give a ‘reasonable’ soloution relatively quickly without taking the item importance into account at all.

The report will include the above items only

To get a Credit

(30+ marks)
The Program must give a ‘reasonable’ soloution relatively quickly which should include some bias to getting higher priority items onto the trucks (note 1=highest importance).

The report will include the above items only for the Pass level run and then the Credit level run

To get a DI

(34+marks)
The Program must give a ‘reasonable’ soloution relatively quickly which should include some bias to getting higher priority items onto the trucks (note 1=highest importance).

The report will also include the above items plus you must repeat the experiments you did in marked tutorial 2, tasks 1, 2 and 3. Predict the score at various generations and comment on whether the findings from marked tutorial 2 are relevant to the GA in this assignment.

To get a HD

(39+marks)
The Program must give a ‘reasonable’ soloution relatively quickly which should include some bias to getting higher priority items onto the trucks (note 1=highest importance). Furthermore the two hospital parts must both be in or out – but there is no point in shipping only one half of the hospital.

The report will also include the above items plus you must repeat the experiments you did in marked tutorial 2, tasks 1, 2 and 3. Predict the score at various generations and comment on whether the findings from marked tutorial 2 are relevant to the GA in this assignment.

Additionally, modify the mutation rate to 25% from the 3% I recommend and demonstrate. Comment on how and why this impacts the answers.

Note – since this is an easier problem the generations may well be 10,20,30,40,60 (not 50, 100, 150, 200, 300)when replicating the experiment done in MT2.
Screen shot from my Pass level example answer

( 2 (