微信客服：xiaoxionga100
微信客服：ITCS521
Python代写-COMP2002
时间：2022-04-10
COMP2002
ARTIFICIAL INTELLIGENCE
20 CREDIT MODULE
ASSESSMENT: 100% Coursework W1: 30% Set Exercises
W2: 70% Report
MODULE LEADER: Dr David Walker
MODULE AIMS
To familiarise with the underlying principles of artificial intelligence.
To expose students to topics such as search and optimization, knowledge representation
and reasoning, and machine learning.
To instil students with an appreciation of the importance of ethical considerations behind
artificial intelligence.
ASSESSED LEARNING OUTCOMES (ALO):
1. Describe and analyse a range of artificial intelligence methods and their applications.
2. Compare artificial intelligence paradigms and evaluate the appropriateness of a particular
paradigm for specific application domains.
3. Choose and apply appropriate artificial intelligence methods to a chosen sample domain.
COMP2002 Artificial Intelligence
Overview
This document contains all the necessary information pertaining to the assessment of COMP2002
Artificial Intelligence. The module is assessed via 100% coursework, across two elements: 30%
Set Exercises and 70% Report.
The sections that follow will detail the assessment tasks that are to be undertaken. The submission
and expected feedback dates are presented in Table 1. All assessments are to be submitted
electronically via the respective DLE module pages before the stated deadlines.
Submission Deadline Feedback
Set Exercises (30%) 19th May 2022, 12pm 15th June 2022
Report (70%) 28th April 2022, 12pm 27th May 2022
Table 1: Assessment Deadlines
All assessments will be introduced in class to provide further clarity over what is expected and how
you can access support and formative feedback prior to submission. Whilst the assessment
information is provided at the start of the module, it is not necessarily expected you will start this
immediately – as you will often not have sufficient understanding of the topic. The module leader will
provide guidance in this respect.
Set Exercises
There are two set exercises for you to complete during the course of the module. Each of these
exercises should take you approximately 8 hours to complete.
COMP2002 Artificial Intelligence
Assessment 1: Set Exercises
Set Exercise 1: worth 40% of element mark
During Lecture 1 you were asked to suggest applications of artificial intelligence. Your task is to select
one of these applications and write a blog post describing it. The topics will be available on the
DLE shortly after the week 1 sessions.
Your blog post must:
Introduce the topic – you should explain the purpose behind the AI. What will it mean we can
now do that we couldn’t before?
Discuss whether the AI has been successful – what has the impact of its introduction been.
Discuss the future of the AI – will it continue to be used? Are there any improvements required
to make it better?
The word limit for your blog post is 700 words.
For information on how to write a good blog post, have a look at the resources posted on the DLE.
You must submit your blog post in PDF form.
Assessment Criteria
A rubric will be used to assess and provide feedback on your submission. This is provided below:
Fail (< 40%) >= 40 >= 50 >= 60 >= 70 >= 80
AI
application
is
introduced
(25%)
Requirements not
met
Vague introduction
with little detail
Some introduction
but some
omissions
Application idea
introduce but more
detail on its
purpose needed
Generally good
introduction – AI
and its purpose
are discussed.
Some minor
omissions.
The AI and its
purpose are fully
introduced. No
omissions.
The AI’s
success is
discussed
(30%)
Requirements not
met
Vague description
of AI’s impact with
little detail. No
discussion.
Some description
of AI’s impact in
some detail. No
discussion.
Good description
of the AI’s impact.
Some analytical
discussion.
The AI’s impact is
well described.
Good analytical
discussion but
lacking detail in a
few places.
Excellent
discussion of the
AI’s impact and
strong analytical
discussion.
The AI’s
future is
discussed
(30%)
Requirements not
met
Vague description
of the AI’s future
with little detail. No
discussion.
Some description
of the AI’s future in
some detail. No
discussion.
Good description
of the AI’s future.
Some analytical
discussion.
The AI’s future is
well described.
Good analytical
discussion but
lacking detail in a
few places.
Excellent
discussion of the
AI’s future and
strong analytical
discussion.
Style and
language
(15%)
Requirements not
met
Poor spelling and
grammar.
Inappropriate
structure.
Poor spelling and
grammar. Some
improvements to
structure required.
Some minor
spelling or
grammatical
errors. Some
improvements to
structure required.
Some minor
spelling or
grammatical
errors. A suitable
structure is
followed.
Excellent. Very few
spelling or
grammatical
errors. A suitable
structure is
followed.
COMP2002 Artificial Intelligence
Assessment 1: Set Exercises
Set Exercise 2: worth 60% of element mark
Your task here is to produce a technical poster about the design of an AI-based assistant to be used
by within a hospital. You must outline the AI components that will go into the assistant, explain how a
user will interact with it.
Your poster post must:
Describe the AI components – how will you use machine learning, evolutionary computation,
knowledge representation and NLP? (You might not use all of these, it’s up to you to decide).
How and why will each be used?
Indicate how the components are assembled into the final system.
Outline the UI – how does the user input requests, and how is the information presented?
Include any ethical considerations.
The poster must be submitted as a PDF file to the DLE and you may use a single page only. A template
has been provided for you on the DLE.
For information on how to write a good scientific poster, have a look at the resources posted on the
DLE.
Assessment Criteria
A rubric will be used to assess and provide feedback on your submission. This is provided below:
Fail (< 40%) >= 40 >= 50 >= 60 >= 70 >= 80
Selection
and
justification
of AI
components
(50%)
Not clear which AI
will be used or
how, no rationale
behind their
selection.
Some of the AIs’
use is described
but there is little
rationale.
The AI to be used
is mostly
described. Some
aspects are
justified but more
detail is needed.
The AI to be used
is described. Most
areas are justified
but justification
could be
strengthened in
places.
The AI to be used
is described and is
mostly justified.
All of the AI to be
used is described
and the rationale
behind the choices
is clear.
UI design
(30%)
Little consideration
to the UI is given,
and there is no
rationale behind
the design.
There is some
consideration to
the UI and some
reasonable
choices have been
made. No
justification.
Reasonable UI
choices have been
made and
justification has
been attempted.
The UI is mostly
well designed. The
design is generally
justified but more
depth is needed.
The UI is well
designed. The
rationale behind
the design is
mostly clear but
could be enhanced
in one or two
areas.
The UI is well
designed and the
rationale behind
the design is clear.
Style and
language
(20%)
Poor spelling and
grammar.
Inappropriate
structure. Poster
content is too brief.
Some spelling and
grammar errors.
Inappropriate
structure. Poster
content is too brief.
Some spelling and
grammar errors.
Minor
improvements to
structure required.
Poster content is
too brief.
Some minor
spelling or
grammatical
errors. Minor
improvements to
structure required.
Some minor
spelling or
grammatical
errors. A suitable
structure is
followed.
Excellent. Very few
spelling or
grammatical
errors. A suitable
structure is
followed.
COMP2002 Artificial Intelligence
Assessment 2: Machine Learning and Optimisation
This assignment contributes 70% of the overall module mark for COMP2002 and is an individual
assignment. You must submit the deliverables to the DLE by the specified submission dates.
The coursework has two parts – one is a machine learning exercise and the second is about
evolutionary computation. You must complete and submit both parts. Each part is worth 50% of
the coursework mark. A Jupyter notebook has been placed on the DLE for you to use. You should
download it and use it to implement the code you need to complete the tasks below.
PART 1 – MACHINE LEARNING
You have been provided with datasets relating to the energy efficiency of buildings. Your task is to
train regression models that predict two indicators of energy efficiency based on 8 numerical inputs.
You must complete the following tasks:
Task 1.1 – Data preparation (10% of total mark)
The first phase of the work requires you to load the data you have been provided with into your Python
program. Before the data can be used to train and test your models you must first prepare it – this
means that the inputs must be normalized. There is no missing data in the dataset.
Task 1.2 – Regression (20% of total mark)
Having prepared the data you must now build a regression tool that can predict new points. Use the
following regression implementations within the scikit-learn package to construct predictors for the
dataset:
Random Forest (sklearn.ensemble.RandomForestRegressor)
Neural Network (sklearn.neural_network.MLPRegressor)
Support Vector Machine (sklearn.svm.SVR)
You must demonstrate that each regressor is capable of providing a prediction for a given input.
Task 1.3 – Assessment of regression (20% of total mark)
The regression models you have used in the previous task must be assessed. To do this you are
required to assess the mean square error rate for each model. You may use the MSE implementation
available in scikit-learn to do this. It is not sufficient to report a single MSE rate. You must use cross
validation to report training and testing results and report these values using a boxplot.
COMP2002 Artificial Intelligence
PART 2 – OPTIMISATION
The second part of this assignment requires you to implement an optimiser to solve a timetabling
problem for a university. You have been provided with a file that describes modules, and lists modules
against which they cannot be scheduled. A module consists of one lecture per week and one or more
lab sessions.
Your task is to optimise a timetable so that session is scheduled once per week in such a way that
timetable constraints are minimised. There are 20 sessions per week. You have one lecture theatre
and two labs available.
Relevant constraints for this assessment are:
A session cannot be scheduled for a time when any of its students or staff are in another
session (concurrence constraints). The sessions for a module that clash are shown in the
data file.
A lab session cannot occur in the week before its corresponding lecture has taken place
(precedence constraints).
You have been provided with a file describing 17 modules, the number of lab sessions, and the
modules that a module cannot conflict with. You must design and implement a fitness function by
taking the number of the concurrence constraints and multiplying them with the number of
precedence constraints. This fitness function should be minimised – the ideal timetable is one with
no constraint violations at all, in which case the function will return 0.
Task 2.1 – Generation of random solutions (10% of total mark)
Your first task is to implement the fitness function. Your code should read the file provided and given
a timetable it should return the quality of the schedule in terms of the constraint violations, as described
above. You should call your function and print out the fitness of a random route.
Task 2.2 – Algorithm implementation (25% of total mark)
You should implement a hillclimber (as described in the lectures) to optimise the problem implemented
in Task 2.1. Your algorithm must have the following features:
It should be possible to use one of two mutation operators – one (session replace) mutates a
solution by placing a session into another slot in the timetable, and the other is a ruin-and-
recreate operator (which generates a completely new solution at random).
At each iteration your single parent solution should be used to generate a single child operator
by using one of the mutation operators (either the swap or the ruin-and-recreate for the entire
run of the algorithm).
At the end of an iteration the algorithm should retain the parent or child that has the best fitness.
At the end of an iteration the best (parent or child) solution’s fitness should be added to a list
of the best fitnesses, which is returned along with the best solution at the end of the
optimisation.
COMP2002 Artificial Intelligence
Task 2.3 – Visualisation of results (15% of total mark)
You should run the algorithm twice for 500 iterations – once for each mutation operator. Repeat this
30 times, so that you get 30 fitness lists for the swap operator and 30 fitness lists for the ruin-and-
recreate operator. Plot the average, maximum and minimum fitness at each iteration for each
operator. You should plot them on the same graph so that they can be compared. You should be able
to see which optimiser is best – state which in the notebook, and say why.
COURSEWORK DELIVERABLES
A Jupyter notebook has been provided on the DLE for you to use for this coursework. You should
implement your code in it, and submit it to the DLE ahead of the deadline specified in the submission
dates earlier in this document. Please indicate which task each section of the notebook refers to
using a Markdown cell.
Please check your submitted files are correct by downloading them again and checking that they work.
You will receive a confirmation receipt by email when your work has been properly submitted
– if you do not receive this email then your work has not been submitted.
ASSESSMENT CRITERIA
Your work will be assessed according to the rubric found in Tables 2 and 3. Your mark for this piece
of coursework will be based on an aggregation of the marks for each category. Marks will be awarded
based on both the demonstration and the report.
Category Fail (< 40%) >= 40% >= 50% >= 60% >= 70%
Data
preparation
(10%)
Data is badly loaded into the
program and there is no real
preparation of the data.
The data is loaded correctly,
though inefficiently, but there
is no preparation.
The data is efficiently loaded
into the program and there is
an attempt at normalisation.
The data is loaded and
normalised, with some
inefficiencies.
The data is loaded and
normalised, and the code to
do so is efficiently written.
/10
Regression
(20%)
Very little or no attempt at
training or testing regression
models. Poor organisation of
code.
There has been an attempt at
training one or more of the
models, but there is no
testing. The organisation of
the code is poor.
All three models are trained,
but there are problems with
the testing. The organisation
of the code is poor.
All three models are trained
and tested correctly. The
code to do so is repetitive
and not well organised.
All three models are trained
and tested correctly. The
code used to do so is well
organised and efficient.
/20
Assessment of
results (20%)
Very little or no attempt at
assessing the regression
results.
There has been an attempt at
using mean absolute error,
but little or no use of cross
validation.
Mean absolute error is used
in combination with cross
validation. The presentation
of results should be
improved.
Mean absolute error is used
in combination with cross
validation, and the results
are presented with a
boxplot. Some inefficiencies
in the code organisation.
Mean absolute error is used
in combination with cross
validation, and the results
are presented with a
boxplot. The code is
efficient and well organised.
/20
Table 2: Feedback Template for Assessment 2 (Machine Learning part)
Category Fail (< 40%) >= 40% >= 50% >= 60% >= 70%
Generation of
random
solutions (10%)
The fitness function is
incorrect. Poor attempt at
loading the data. Generation
of random solutions does not
work.
The data is loaded. There
has been an attempt at the
fitness function, but it is
incorrect. Generation of
random solutions does not
work.
The data is loaded and the
fitness function is close to
correct. Generation of
random solutions does not
work.
The data is loaded and the
fitness function is close to
correct. Solution generation
is close to working. Code is
inefficiently designed.
The data is loaded and the
fitness function is correct.
Random solutions are
generated correctly. The
code to do so is efficiently
structured.
/10
Algorithm
implementation
(25%)
Mutation operators are mostly
incorrect or missing. The
algorithm is incorrect –
selection is missing or not
working and the fitness
archive is missing.
One mutation operator has
been attempted. Some of the
algorithm is present, but it is
incomplete or incorrect. The
fitness archive has been
attempted.
Two mutation operators have
been attempted but are
incomplete or incorrect. The
algorithm is partially correct
and the fitness archive works
correctly.
The mutation operators
work and the algorithm is
mostly implemented
correctly. The fitness
archive stores fitnesses
correctly.
The mutation operators and
algorithm have been
implemented correctly. The
fitness archive stores
fitnesses correctly. The
code is efficiently structured
and well organised.
/25
Visualisation of
results (15%)
No attempt at producing a
plot, or the plot is incorrect.
Experiments are not run for
the correct number of
repeats.
A graph shows the partial
results required. The
experiment was run
incorrectly.
A graph shows the correct
results required but the
experiment was run
incorrectly.
A graph shows the correct
results and the correct
experimental setup has
been followed. No analysis
of the results.
A graph shows the correct
results and the correct
experimental setup has
been followed. The best
mutation operator is
identified and justified.
/15
Table 3: Feedback Template for Assessment 2 (Evolutionary Computation part)
COMP2002 Artificial Intelligence
General Guidance
Extenuating Circumstances
There may be a time during this module where you experience a serious situation which has a
significant impact on your ability to complete the assessments. The definition of these can be found
in the University Policy on Extenuating Circumstances here:
https://www.plymouth.ac.uk/uploads/production/document/path/15/15317/Extenuating_Circumstan
ces_Policy_and_Procedures.pdf
Plagiarism
All of your work must be of your own words. You must use references for your sources, however
you acquire them. Where you wish to use quotations, these must be a very minor part of your overall
work.
To copy another person’s work is viewed as plagiarism and is not allowed. Any issues of plagiarism
and any form of academic dishonesty are treated very seriously. All your work must be your own
and other sources must be identified as being theirs, not yours. The copying of another persons’
work could result in a penalty being invoked.
Further information on plagiarism policy can be found here:
Plagiarism: https://www.plymouth.ac.uk/student-life/your-studies/essential-
information/regulations/plagiarism
Examination Offences: https://www.plymouth.ac.uk/student-life/your-studies/essential-
information/exams/exam-rules-and-regulations/examination-offences
Turnitin (http://www.turnitinuk.com/) is an Internet-based 'originality checking tool' which allows
documents to be compared with content on the Internet, in journals and in an archive of previously
submitted works. It can help to detect unintentional or deliberate plagiarism.
It is a formative tool that makes it easy for students to review their citations and referencing as an
aid to learning good academic practice. Turnitin produces an ‘originality report’ to help guide you.
To learn more about Turnitin go to:
https://guides.turnitin.com/01_Manuals_and_Guides/Student/Student_User_Manual
ARTIFICIAL INTELLIGENCE
20 CREDIT MODULE
ASSESSMENT: 100% Coursework W1: 30% Set Exercises
W2: 70% Report
MODULE LEADER: Dr David Walker
MODULE AIMS
To familiarise with the underlying principles of artificial intelligence.
To expose students to topics such as search and optimization, knowledge representation
and reasoning, and machine learning.
To instil students with an appreciation of the importance of ethical considerations behind
artificial intelligence.
ASSESSED LEARNING OUTCOMES (ALO):
1. Describe and analyse a range of artificial intelligence methods and their applications.
2. Compare artificial intelligence paradigms and evaluate the appropriateness of a particular
paradigm for specific application domains.
3. Choose and apply appropriate artificial intelligence methods to a chosen sample domain.
COMP2002 Artificial Intelligence
Overview
This document contains all the necessary information pertaining to the assessment of COMP2002
Artificial Intelligence. The module is assessed via 100% coursework, across two elements: 30%
Set Exercises and 70% Report.
The sections that follow will detail the assessment tasks that are to be undertaken. The submission
and expected feedback dates are presented in Table 1. All assessments are to be submitted
electronically via the respective DLE module pages before the stated deadlines.
Submission Deadline Feedback
Set Exercises (30%) 19th May 2022, 12pm 15th June 2022
Report (70%) 28th April 2022, 12pm 27th May 2022
Table 1: Assessment Deadlines
All assessments will be introduced in class to provide further clarity over what is expected and how
you can access support and formative feedback prior to submission. Whilst the assessment
information is provided at the start of the module, it is not necessarily expected you will start this
immediately – as you will often not have sufficient understanding of the topic. The module leader will
provide guidance in this respect.
Set Exercises
There are two set exercises for you to complete during the course of the module. Each of these
exercises should take you approximately 8 hours to complete.
COMP2002 Artificial Intelligence
Assessment 1: Set Exercises
Set Exercise 1: worth 40% of element mark
During Lecture 1 you were asked to suggest applications of artificial intelligence. Your task is to select
one of these applications and write a blog post describing it. The topics will be available on the
DLE shortly after the week 1 sessions.
Your blog post must:
Introduce the topic – you should explain the purpose behind the AI. What will it mean we can
now do that we couldn’t before?
Discuss whether the AI has been successful – what has the impact of its introduction been.
Discuss the future of the AI – will it continue to be used? Are there any improvements required
to make it better?
The word limit for your blog post is 700 words.
For information on how to write a good blog post, have a look at the resources posted on the DLE.
You must submit your blog post in PDF form.
Assessment Criteria
A rubric will be used to assess and provide feedback on your submission. This is provided below:
Fail (< 40%) >= 40 >= 50 >= 60 >= 70 >= 80
AI
application
is
introduced
(25%)
Requirements not
met
Vague introduction
with little detail
Some introduction
but some
omissions
Application idea
introduce but more
detail on its
purpose needed
Generally good
introduction – AI
and its purpose
are discussed.
Some minor
omissions.
The AI and its
purpose are fully
introduced. No
omissions.
The AI’s
success is
discussed
(30%)
Requirements not
met
Vague description
of AI’s impact with
little detail. No
discussion.
Some description
of AI’s impact in
some detail. No
discussion.
Good description
of the AI’s impact.
Some analytical
discussion.
The AI’s impact is
well described.
Good analytical
discussion but
lacking detail in a
few places.
Excellent
discussion of the
AI’s impact and
strong analytical
discussion.
The AI’s
future is
discussed
(30%)
Requirements not
met
Vague description
of the AI’s future
with little detail. No
discussion.
Some description
of the AI’s future in
some detail. No
discussion.
Good description
of the AI’s future.
Some analytical
discussion.
The AI’s future is
well described.
Good analytical
discussion but
lacking detail in a
few places.
Excellent
discussion of the
AI’s future and
strong analytical
discussion.
Style and
language
(15%)
Requirements not
met
Poor spelling and
grammar.
Inappropriate
structure.
Poor spelling and
grammar. Some
improvements to
structure required.
Some minor
spelling or
grammatical
errors. Some
improvements to
structure required.
Some minor
spelling or
grammatical
errors. A suitable
structure is
followed.
Excellent. Very few
spelling or
grammatical
errors. A suitable
structure is
followed.
COMP2002 Artificial Intelligence
Assessment 1: Set Exercises
Set Exercise 2: worth 60% of element mark
Your task here is to produce a technical poster about the design of an AI-based assistant to be used
by within a hospital. You must outline the AI components that will go into the assistant, explain how a
user will interact with it.
Your poster post must:
Describe the AI components – how will you use machine learning, evolutionary computation,
knowledge representation and NLP? (You might not use all of these, it’s up to you to decide).
How and why will each be used?
Indicate how the components are assembled into the final system.
Outline the UI – how does the user input requests, and how is the information presented?
Include any ethical considerations.
The poster must be submitted as a PDF file to the DLE and you may use a single page only. A template
has been provided for you on the DLE.
For information on how to write a good scientific poster, have a look at the resources posted on the
DLE.
Assessment Criteria
A rubric will be used to assess and provide feedback on your submission. This is provided below:
Fail (< 40%) >= 40 >= 50 >= 60 >= 70 >= 80
Selection
and
justification
of AI
components
(50%)
Not clear which AI
will be used or
how, no rationale
behind their
selection.
Some of the AIs’
use is described
but there is little
rationale.
The AI to be used
is mostly
described. Some
aspects are
justified but more
detail is needed.
The AI to be used
is described. Most
areas are justified
but justification
could be
strengthened in
places.
The AI to be used
is described and is
mostly justified.
All of the AI to be
used is described
and the rationale
behind the choices
is clear.
UI design
(30%)
Little consideration
to the UI is given,
and there is no
rationale behind
the design.
There is some
consideration to
the UI and some
reasonable
choices have been
made. No
justification.
Reasonable UI
choices have been
made and
justification has
been attempted.
The UI is mostly
well designed. The
design is generally
justified but more
depth is needed.
The UI is well
designed. The
rationale behind
the design is
mostly clear but
could be enhanced
in one or two
areas.
The UI is well
designed and the
rationale behind
the design is clear.
Style and
language
(20%)
Poor spelling and
grammar.
Inappropriate
structure. Poster
content is too brief.
Some spelling and
grammar errors.
Inappropriate
structure. Poster
content is too brief.
Some spelling and
grammar errors.
Minor
improvements to
structure required.
Poster content is
too brief.
Some minor
spelling or
grammatical
errors. Minor
improvements to
structure required.
Some minor
spelling or
grammatical
errors. A suitable
structure is
followed.
Excellent. Very few
spelling or
grammatical
errors. A suitable
structure is
followed.
COMP2002 Artificial Intelligence
Assessment 2: Machine Learning and Optimisation
This assignment contributes 70% of the overall module mark for COMP2002 and is an individual
assignment. You must submit the deliverables to the DLE by the specified submission dates.
The coursework has two parts – one is a machine learning exercise and the second is about
evolutionary computation. You must complete and submit both parts. Each part is worth 50% of
the coursework mark. A Jupyter notebook has been placed on the DLE for you to use. You should
download it and use it to implement the code you need to complete the tasks below.
PART 1 – MACHINE LEARNING
You have been provided with datasets relating to the energy efficiency of buildings. Your task is to
train regression models that predict two indicators of energy efficiency based on 8 numerical inputs.
You must complete the following tasks:
Task 1.1 – Data preparation (10% of total mark)
The first phase of the work requires you to load the data you have been provided with into your Python
program. Before the data can be used to train and test your models you must first prepare it – this
means that the inputs must be normalized. There is no missing data in the dataset.
Task 1.2 – Regression (20% of total mark)
Having prepared the data you must now build a regression tool that can predict new points. Use the
following regression implementations within the scikit-learn package to construct predictors for the
dataset:
Random Forest (sklearn.ensemble.RandomForestRegressor)
Neural Network (sklearn.neural_network.MLPRegressor)
Support Vector Machine (sklearn.svm.SVR)
You must demonstrate that each regressor is capable of providing a prediction for a given input.
Task 1.3 – Assessment of regression (20% of total mark)
The regression models you have used in the previous task must be assessed. To do this you are
required to assess the mean square error rate for each model. You may use the MSE implementation
available in scikit-learn to do this. It is not sufficient to report a single MSE rate. You must use cross
validation to report training and testing results and report these values using a boxplot.
COMP2002 Artificial Intelligence
PART 2 – OPTIMISATION
The second part of this assignment requires you to implement an optimiser to solve a timetabling
problem for a university. You have been provided with a file that describes modules, and lists modules
against which they cannot be scheduled. A module consists of one lecture per week and one or more
lab sessions.
Your task is to optimise a timetable so that session is scheduled once per week in such a way that
timetable constraints are minimised. There are 20 sessions per week. You have one lecture theatre
and two labs available.
Relevant constraints for this assessment are:
A session cannot be scheduled for a time when any of its students or staff are in another
session (concurrence constraints). The sessions for a module that clash are shown in the
data file.
A lab session cannot occur in the week before its corresponding lecture has taken place
(precedence constraints).
You have been provided with a file describing 17 modules, the number of lab sessions, and the
modules that a module cannot conflict with. You must design and implement a fitness function by
taking the number of the concurrence constraints and multiplying them with the number of
precedence constraints. This fitness function should be minimised – the ideal timetable is one with
no constraint violations at all, in which case the function will return 0.
Task 2.1 – Generation of random solutions (10% of total mark)
Your first task is to implement the fitness function. Your code should read the file provided and given
a timetable it should return the quality of the schedule in terms of the constraint violations, as described
above. You should call your function and print out the fitness of a random route.
Task 2.2 – Algorithm implementation (25% of total mark)
You should implement a hillclimber (as described in the lectures) to optimise the problem implemented
in Task 2.1. Your algorithm must have the following features:
It should be possible to use one of two mutation operators – one (session replace) mutates a
solution by placing a session into another slot in the timetable, and the other is a ruin-and-
recreate operator (which generates a completely new solution at random).
At each iteration your single parent solution should be used to generate a single child operator
by using one of the mutation operators (either the swap or the ruin-and-recreate for the entire
run of the algorithm).
At the end of an iteration the algorithm should retain the parent or child that has the best fitness.
At the end of an iteration the best (parent or child) solution’s fitness should be added to a list
of the best fitnesses, which is returned along with the best solution at the end of the
optimisation.
COMP2002 Artificial Intelligence
Task 2.3 – Visualisation of results (15% of total mark)
You should run the algorithm twice for 500 iterations – once for each mutation operator. Repeat this
30 times, so that you get 30 fitness lists for the swap operator and 30 fitness lists for the ruin-and-
recreate operator. Plot the average, maximum and minimum fitness at each iteration for each
operator. You should plot them on the same graph so that they can be compared. You should be able
to see which optimiser is best – state which in the notebook, and say why.
COURSEWORK DELIVERABLES
A Jupyter notebook has been provided on the DLE for you to use for this coursework. You should
implement your code in it, and submit it to the DLE ahead of the deadline specified in the submission
dates earlier in this document. Please indicate which task each section of the notebook refers to
using a Markdown cell.
Please check your submitted files are correct by downloading them again and checking that they work.
You will receive a confirmation receipt by email when your work has been properly submitted
– if you do not receive this email then your work has not been submitted.
ASSESSMENT CRITERIA
Your work will be assessed according to the rubric found in Tables 2 and 3. Your mark for this piece
of coursework will be based on an aggregation of the marks for each category. Marks will be awarded
based on both the demonstration and the report.
Category Fail (< 40%) >= 40% >= 50% >= 60% >= 70%
Data
preparation
(10%)
Data is badly loaded into the
program and there is no real
preparation of the data.
The data is loaded correctly,
though inefficiently, but there
is no preparation.
The data is efficiently loaded
into the program and there is
an attempt at normalisation.
The data is loaded and
normalised, with some
inefficiencies.
The data is loaded and
normalised, and the code to
do so is efficiently written.
/10
Regression
(20%)
Very little or no attempt at
training or testing regression
models. Poor organisation of
code.
There has been an attempt at
training one or more of the
models, but there is no
testing. The organisation of
the code is poor.
All three models are trained,
but there are problems with
the testing. The organisation
of the code is poor.
All three models are trained
and tested correctly. The
code to do so is repetitive
and not well organised.
All three models are trained
and tested correctly. The
code used to do so is well
organised and efficient.
/20
Assessment of
results (20%)
Very little or no attempt at
assessing the regression
results.
There has been an attempt at
using mean absolute error,
but little or no use of cross
validation.
Mean absolute error is used
in combination with cross
validation. The presentation
of results should be
improved.
Mean absolute error is used
in combination with cross
validation, and the results
are presented with a
boxplot. Some inefficiencies
in the code organisation.
Mean absolute error is used
in combination with cross
validation, and the results
are presented with a
boxplot. The code is
efficient and well organised.
/20
Table 2: Feedback Template for Assessment 2 (Machine Learning part)
Category Fail (< 40%) >= 40% >= 50% >= 60% >= 70%
Generation of
random
solutions (10%)
The fitness function is
incorrect. Poor attempt at
loading the data. Generation
of random solutions does not
work.
The data is loaded. There
has been an attempt at the
fitness function, but it is
incorrect. Generation of
random solutions does not
work.
The data is loaded and the
fitness function is close to
correct. Generation of
random solutions does not
work.
The data is loaded and the
fitness function is close to
correct. Solution generation
is close to working. Code is
inefficiently designed.
The data is loaded and the
fitness function is correct.
Random solutions are
generated correctly. The
code to do so is efficiently
structured.
/10
Algorithm
implementation
(25%)
Mutation operators are mostly
incorrect or missing. The
algorithm is incorrect –
selection is missing or not
working and the fitness
archive is missing.
One mutation operator has
been attempted. Some of the
algorithm is present, but it is
incomplete or incorrect. The
fitness archive has been
attempted.
Two mutation operators have
been attempted but are
incomplete or incorrect. The
algorithm is partially correct
and the fitness archive works
correctly.
The mutation operators
work and the algorithm is
mostly implemented
correctly. The fitness
archive stores fitnesses
correctly.
The mutation operators and
algorithm have been
implemented correctly. The
fitness archive stores
fitnesses correctly. The
code is efficiently structured
and well organised.
/25
Visualisation of
results (15%)
No attempt at producing a
plot, or the plot is incorrect.
Experiments are not run for
the correct number of
repeats.
A graph shows the partial
results required. The
experiment was run
incorrectly.
A graph shows the correct
results required but the
experiment was run
incorrectly.
A graph shows the correct
results and the correct
experimental setup has
been followed. No analysis
of the results.
A graph shows the correct
results and the correct
experimental setup has
been followed. The best
mutation operator is
identified and justified.
/15
Table 3: Feedback Template for Assessment 2 (Evolutionary Computation part)
COMP2002 Artificial Intelligence
General Guidance
Extenuating Circumstances
There may be a time during this module where you experience a serious situation which has a
significant impact on your ability to complete the assessments. The definition of these can be found
in the University Policy on Extenuating Circumstances here:
https://www.plymouth.ac.uk/uploads/production/document/path/15/15317/Extenuating_Circumstan
ces_Policy_and_Procedures.pdf
Plagiarism
All of your work must be of your own words. You must use references for your sources, however
you acquire them. Where you wish to use quotations, these must be a very minor part of your overall
work.
To copy another person’s work is viewed as plagiarism and is not allowed. Any issues of plagiarism
and any form of academic dishonesty are treated very seriously. All your work must be your own
and other sources must be identified as being theirs, not yours. The copying of another persons’
work could result in a penalty being invoked.
Further information on plagiarism policy can be found here:
Plagiarism: https://www.plymouth.ac.uk/student-life/your-studies/essential-
information/regulations/plagiarism
Examination Offences: https://www.plymouth.ac.uk/student-life/your-studies/essential-
information/exams/exam-rules-and-regulations/examination-offences
Turnitin (http://www.turnitinuk.com/) is an Internet-based 'originality checking tool' which allows
documents to be compared with content on the Internet, in journals and in an archive of previously
submitted works. It can help to detect unintentional or deliberate plagiarism.
It is a formative tool that makes it easy for students to review their citations and referencing as an
aid to learning good academic practice. Turnitin produces an ‘originality report’ to help guide you.
To learn more about Turnitin go to:
https://guides.turnitin.com/01_Manuals_and_Guides/Student/Student_User_Manual
