6G5Z0032 Machine Learning – 1CWK50

Assessment Cover Sheet 2025/26

Module code and title 6G5Z0032 Machine Learning

Assessment set by John Darby

Assessment ID 1CWK50

Assessment weighting 50%

Assessment title Coursework

Type Individual

Hand-in deadline 23rd March 2026

Via Moodle, electronic .zip file (or link to your OneDrive

Hand-in format and mechanism

for large submissions >100MB)

Learning outcomes being assessed

LO1: Explain theoretical and practical aspects of popular, state-of-the-art supervised and unsupervised

machine learning techniques and models.

LO2: Select, implement, apply, optimise and evaluate suitable machine learning algorithms for given domains

to create a range of models from data.

Note: it is your responsibility to make sure that your work is complete and available for marking by the

deadline. Make sure that you have followed the submission instructions carefully, and your work is submitted

in the correct format, using the correct hand-in mechanism (e.g., Moodle upload). If submitting via Moodle,

you are advised to check your work after upload, to make sure it has uploaded properly. If submitting via

OneDrive, ensure that your tutors have access to the work. Do not alter your work after the deadline. You

should make at least one full backup copy of your work.

Penalties for late submission

The timeliness of submissions is strictly monitored and enforced.

All coursework has a late submission window of 7 calendar days, but any work submitted within the late

window will be capped at 40%, unless you have an agreed extension. Work submitted after the 7-day late

window will be capped at zero unless you have an agreed extension. See ‘Assessment Mitigation’ below for

further information on extensions.

Please note that individual tutors are unable to grant any extensions to assessments.

Assessment Mitigation

If there is a valid reason why you are unable to submit your assessment by the deadline you may apply for

Assessment Mitigation. There are two types of mitigation you can apply for via the module area on Moodle (in

the ‘Assessments’ block on the right-hand side of the page):

• Non-evidenced extension: does not require you to submit evidence. It allows you to add a short extension

to a deadline. This is not available for event-based assessments such as in-class tests, presentations,

interviews, etc. You can apply for this extension during the assessment weeks, and the request must be

made before the submission deadline. For this assessment, the non-evidenced extension is 2 days.

• Evidenced extension: requires you to provide independent evidence of a situation which has impacted

you. Allows you to apply for a longer extension and is available for event-based assessment such as in-

class test, presentations, interviews, etc. For event-based assessments, the normal outcome is that the

assessment will be deferred to the summer reassessment period.

Further information about Assessment Mitigation is available on the dedicated Assessments page.

Personal Learning Plans (PLP)

If you have a Personal Learning Plan (PLP) which states you can negotiate an extended deadline, submit an

evidenced extension request on the Moodle site for the module.

Plagiarism

Plagiarism is the unacknowledged representation of another person’s work, or use of their ideas, as one’s own.

Manchester Metropolitan University takes care to detect plagiarism, employs plagiarism detection software,

and imposes severe penalties, as outlined in the Student Code of Conduct and Academic integrity and

misconduct. Poor referencing or submitting the wrong assignment may still be treated as plagiarism. If in

doubt, seek advice from your tutor.

As part of a plagiarism check, you may be asked to attend a meeting with the Module Leader, or another

member of the module delivery team, where you will be asked to explain your work (e.g. explain the code

in a programming assignment). If you are called to one of these meetings, it is very important that you

attend.

Use of generative AI

The use of generative AI is permitted in this assessment, so long as it is used in accordance with the

instructions provided in the ‘Are you allowed to use AI in assessments?’ section of the AI Literacy Rise Study

Pack. All submitted work must be your own original content.

If you are unable to upload your work to Moodle

If you have problems submitting your work through Moodle, you can raise a ticket with the Assessment

Management Team using the Assist Portal. This must be done before the published deadline, else your work

will be logged as a late submission. Alternatively, you can save your work into a single zip folder then upload the

zip folder to your university OneDrive and submit a Word document to Moodle which includes a link to the

folder. It is your responsibility to make sure you share the OneDrive folder with the Module Leader, or it

will not be possible to mark your work.

Assessment Regulations

For further information see the Undergraduate Assessment Regulations on the Assessments and Results

information pages

Formative feedback: You can discuss your assignment work with your tutor in any of the on-

campus sessions, or by getting in contact with them directly (see the

“module information” page on Moodle for contact details).

Summative feedback: A final summative mark and a completed marksheet containing

highlighted criteria and feedback will be made available via Moodle. See

the attached specification for more details.

1CWK50: Visual explanations in PowerPoint

1. Challenge

In the early weeks of this module, we learn about some different supervised learning models (approximately 1 model

per week). The challenge in this assignment is to produce an explanation of how these models work, using PowerPoint.

The twist to the challenge is that your PowerPoint files cannot contain any words.1 You will need to rely solely on

visuals to convey, in detail, how each model works. Furthermore, all your visuals must be created either with

PowerPoint’s built-in tools (e.g., Shapes, Icons, and freehand drawing) or from images you have personally authored

(e.g., figures generated using Python code in labs). No images created by others may be used.

2. Specifics

You should explain each of the models listed below. They are all supervised learning models for classification problems,

capable of discovering patterns between groups of input values and single output values. Each model can be thought

of as a class (in the computer science/programming sense) with a .fit(inputs, outputs) method for learning patterns,

and a .predict(inputs) method for using previously learned patterns to generate new output predictions based on new

inputs. Your task is to explain how both these methods work, for each model. You should assume that the inputs are

always numerical values, and that any necessary data pre-processing steps have already taken place (i.e., you don’t

need to explain extra steps like cleaning, standardisation, etc.).

1. A k-nearest neighbours classifier

2. A naïve Bayes classifier

3. A decision tree classifier

4. An ensemble classifier

While the first 3 models are very specific, the final one is more open-ended, and a chance to go as far/deep as you

wish in your explanation. Credit is available for being ambitious – see also Section 3.

If you want to check any details of model behaviour, then you should refer to the model implementations available in

the Scikit-learn Python package. (This is the machine learning package we use in labs.)

3. Requirements

We will consider the following requirements when marking your submission:

1. Accuracy: the technical details that you cover in the slides must be accurate;

2. Completeness: the technical details that you cover in the slides must be complete, without any steps missing;

3. Clarity: your approach to conveying the technical details that you cover must be as precise and unambiguous

as possible, minimising the potential for any misunderstandings;

4. Consistency: as well as being clear at each individual step in your explanation (e.g., within a single slide), you

should also aim to ensure the consistency of your explanation between different steps (e.g., between

individual slides);

5. Accessibility: your slides should minimise demands on the reader and make it as easy as possible for anyone

to understand the models in full – achieving this using only visuals will require careful and creative thinking,

including thinking from the perspective of someone who doesn’t yet understand the model (even though you

may understand the model really well yourself!);

6. Ambition: your slides should be ambitious in their scope – striving to give the reader the fullest possible picture

of how each model works. For example:

a. Most of the models have interesting hyperparameters which give a user some control over exactly

what happens inside the fit() method – could you extend your explanations to include useful insight

into one or more hyperparameters?

1 Numbers and mathematical notation are an exception but see also the Accessibility requirement in Section 3.

b. Approximate explanations of the fit() and predict() methods (e.g., via freehand “sketching”) are good,

but could you use real data and real parameter/hyperparameter values to ensure that ideas are

communicated in a concrete and internally consistent way?

c. Most of the models have relatively simple 1D or 2D interpretations (when there are just 1 or 2 inputs),

but could you give the reader insight in to how they work with higher-dimensional data (that has more

than 2 inputs)?

7. Authenticity: overall, the submission, and your engagement with the module, suggests that you have worked

hard to develop relevant skills/understanding, and that you have retained ultimate authorship of the work

throughout (see also Section 7).

4. Resources

Relevant foundational theory and practical skills are covered in the learning materials, labs and lectures for this

module. However, you are also expected to bring your own existing skills, creativity and ingenuity to bear.

You can find a simple template PowerPoint (.pptx) file below this assignment specification on Moodle.

Recall that no images made by others may be used.

5. Support

Help is always available in your weekly labs, or by contacting your tutor directly (contact details are included on the

“Module Information” page on Moodle). Bear in mind there are some limits on the help that can be given without

interfering with the assessment of the learning objectives.

6. How to submit your work

Once you’re ready to submit, you should compress your PowerPoint (.pptx) file into a single .zip archive file (no other

compression formats are accepted), and upload it to Moodle.

If your .zip is too big for Moodle you can upload it to your OneDrive, and submit to Moodle a (compressed) document

containing a link to the OneDrive copy instead. (See also the standard cover sheet instructions at the start of this

document.)

7. Authorship (and use of generative AI)

The work you submit for this assignment is expected to represent your own efforts to explain the different models. Be

careful when giving/accepting help to/from friends, and make sure that any other sources of information or ideas are

properly cited and referenced within your PowerPoint file.2 This includes the source files for any self-authored images.

provided in the ‘Are you allowed to use AI in assessments?’ section of the AI Literacy Rise Study Pack. Essentially: AI

can be used as a tool to help you learn, but not as a tool to do your assessment for you.

8. Feedback

You can discuss your assignment work with your tutor in any of the on-campus sessions, or by getting in contact with

them directly (see the “Module Information” page on Moodle for contact details and office hours).

Your final summative feedback sheet for this assignment will consist of a copy of the marking grid set out in this

specification document (see Section 9) with the x’s shifted to indicate your level of performance across each of the

requirements set out in Section 3, plus written feedback from your tutor, and a final stepped3 mark for the overall

assignment out of 100. The feedback sheet will be returned to you via Moodle.

2 You should use Cite Them Right Harvard for the formatting of any citations/references in your submission. Guidance is

available via the MMU library webpages here.

3 Only certain percentages are awarded within each classification band – usually #2. #5 and #8 (e.g., 62, 65 and 68).

9. Marking criteria

Accurac Complet Clarity Accessib Consiste Ambitio Authent

Mark Indicative language

-y -eness -ility -ncy -n -icity

100

98

95

92

Persuasive, sophisticated,

88

original, reflective, ambitious,

F

ir meticulous, critical,

s 85

t

convincing, unexpected

82

78

75

72

68

Fluent, thorough, analytical,

2 65

:1 precise, rigorous

62

58

Clear, confident, consistent,

2 55 thoughtful, accurate, careful,

:2

congruent, coherent

52

48

Satisfactory, sufficient,

45

3

rd adequate, descriptive

42

x x x x x x x

38 Incomplete, inadequate,

C

o

n inconsistent, derivative,

fa d o 35 contradictory, superficial,

il n

a

b 32 irrelevant, limited

le

28

25

22

18 Erroneous / wrong, missing,

extremely limited,

15

inappropriate, insufficient,

a incoherent, unstructured;

il 12

absent/none, lacking,

8 formless, detrimental

5

2

0

📩 Want the Full Solution?

Contact us via WhatsApp or email with this assignment title and we'll send you the complete solution — including step-by-step explanation, references, and Turnitin report.