You are on page 1of 3

College of Engineering & Physical Sciences

Assignment Brief

CS4755_P2_Mathematics for AI Written Report

Dr Mohammed Hadi Dr Liam Escott


m.hadi2@aston.ac.uk l.escott@aston.ac.uk

Assignment Brief/ Coursework Content:

In this coursework, students work on individual topics. They select a suitable problem or
application from the context of AI and analyse it – focusing on the mathematical foundations
involved in solving the problem. The topic can relate to the student’s own MSc project, or some
practical application within the context of AI.
The analysis should be presented as a report and should contain the following contents:
- Introduction to the topic: presentation of the use cases or applications.
- Brief review of method(s) that can be applied to solve the chosen use case.
- Analysis of the method(s) focusing particularly on their mathematical foundation(s) and
how they relate to the topic(s) covered in the lecture(s).
- A numerical example that demonstrates the key mathematical foundations of the
described method(s). Programming code can be suitable where it is very difficult to
provide a detailed numerical example. In any case, students should provide the
corresponding mathematical formulations.
- Discussion of the underlying mathematical assumptions, potential strengths and
weaknesses of the method(s) described. Reflection and analysis of how the described
method(s) potentially impact the considered use cases.

Report Style and Organisation:


General guidance: Universal standards of academic writing apply in terms of style of writing
(tone and diction) as well as structure, contents, and the presentation of evidence-based
arguments. The report should have an appropriate title that describes the specifics of the
topic in a concise and accurate way. An abstract is not required. The structure and, in
particular, the length of each section should be chosen appropriately based on their
relevance.

Word limit: The word limit is set to 1,500 words, and an extra 10% is tolerated (to the
absolute maximum of 1,650 words). The following parts of the assignment do not contribute
towards the word count:
(1) The list of references added in a section at the end of the report,
(2) Figures, equations, or tables.
(3) Appendices containing code.
All other submitted text should be assumed to count towards the word limit.

Figures, Tables, Equations: Figures, tables, and equations should be numbered and
referred to by their number from the main text. Figures and tables must have a descriptive
caption. Equations should be formatted using proper mathematical fonts and all variable
names should be explained.

Referencing: Use IEEE style for references, i.e., when using a statement from a certain
source within the text, refer to the source by its number (e.g. [1]) and provide a section titled
“References” at the end of the body of the report (before a potential appendix) containing the
full information about all used references. E.g.:
[1] J. W. Hall, “Adaptive selection of U.S. stocks with neural nets,” in Trading on the Edge: Neural, Genetic,
and Fuzzy Systems for Chaotic Financial Markets. New York: Wiley, 1994

Use at least five references.

Style and formatting: Use single-column page format. Pages and sections must be
numbered. The document should be set in a professional font type with appropriate size,
e.g., 12pt Arial.

Descriptive details of Assignment:


• Document Title e.g., Firstname_Lastname_CS4755_P2_Report
• Submission Format is a single PDF or .docx file
• Word Limit: 1,500, + 10% is tolerated.

Recommended reading/ online sources:

• K. P. Murphy, 2012, Machine Learning: A Probabilistic Perspective MIT Press


• Aston LDC Academic Writing. Check available support by following this link.
• Documents available on Blackboard.

Key Dates:

06/02/2024 Coursework set.


26/02/2024 You may submit your selected topic by this date, although you
can submit earlier.
(Online via Blackboard > Assessment Submission)
By 5:00 PM Report submission date and time
23/04/2024

Submission Details:
• Electronic submission (online on Blackboard) via Turnitin.

Marking Rubric:

The report will be marked with respect to different criteria including Technical Quality, Writing
Quality, Presentation, Discussion, Literature Review and Referencing, as follows:

0-24 - Fragmentary or almost wholly ineffective activity.

25-49 - Work that provides evidence of relevant knowledge and skills but deployed on such a
limited scale or in such an ineffective manner that the threshold standard for the module has
not been achieved. For example: use case and practical relevance remains unclear; literature
review is too brief and includes misunderstandings or misinterpretations; there are some signs
of task comprehension but only on a superficial level; methods benefits or drawbacks were
not discussed; figures, tables and equations are of questionable quality or relevance; an
overall poor report structure and spelling/grammar.

50-59 - The work is motivated in general and a range of expected activities have been
undertaken, but at a minimal standard of competence and understanding leading to a barely
acceptable outcome: literature review demonstrates general understanding of the subject
although missing some elements; figures, tables and equations generally support the line of
thought but are of varying quality; formatting is consistent and all references have been
provided; the structure is reasonable and no important elements have been missed (related
to the mathematical model evaluation); some sections may require re-reading but the line of
thought can generally be followed; the contribution is mostly clear and supported by numerical
examples; an attempt was made to compare methods empirically.

60-69 - Expected activities (applying mathematical concepts that are used to solve AI
problems and model evaluation) have been undertaken in a manner that shows a clear grasp
of what is required and leads to a broadly acceptable outcome that has undergone evaluation.
The report is the result of systematic work; the review covers most important method(s) with
only minor gaps or misunderstandings; appropriate use and formatting of references, although
of varying quality; the work demonstrates good knowledge and understanding of the subject;
equations, figures and tables support the line of thought and are properly formatted; structure
is generally sound and no parts are missed; the work is generally consistent and the
mathematical foundations are clearly demonstrated using numerical examples; advantages
and drawbacks are being discussed on a general level.

70-79 - Work producing a sound outcome: the report is the result of systematic work, with
observations supported by suitable references and logical reasoning. The review is insightful
and uses references where appropriate, providing mostly primary sources with scientific
credibility; the analysis is consistent and clearly demonstrates understanding of the
mathematical foundations; the benefits and drawbacks of their application are being discussed
in relation to the use case; grammar/spelling is almost always correct; the writing is mostly
clear and concise; figures, tables and equations are of good quality and support the line of
thought where suitable.

80-89 - Work distinguished by evidence of a deeper level of understanding, insight and


innovation, and with very few (if any) flaws. The use case is well chosen and clearly presented;
the review provides insights into relevant methods; the analysis demonstrates deep
understanding of the relevant mathematical foundations, and a well-chosen numerical
example supports that. The writing is clear and concise, with no errors in grammar/spelling;
the structure is fluid and gives clear priority to the most important parts of the work. Figures,
tables and equations are appealing and uniformly formatted, providing additional value over
the writing. The line of arguments is original and convincing. Benefits and drawbacks of the
methods are clearly pointed out, as well as potential assumptions that are made for being able
to apply the method.

90+ - Exceptional work approximating a professional standard that a member of academic


staff would be satisfied to have produced themselves. The choice and presentation of methods
in the review demonstrates deep understanding of the subject; references are of high scientific
quality; formatting and layout is flawless and appealing. Equations, tables and figures are
insightful and add value, while the writing is clear and concise, with even complex subjects
being described in an easily comprehensible manner. The work is focused, clearly points out
the relevant mathematical foundations, reflects assumptions, benefits and drawbacks that
apply to the particular use case that has been chosen. The work demonstrates excellent
understanding as well as critical thinking.

You might also like