CRN 61047 / 61048 Machine Learning and Data Mining (MLDM) Level 7 Assessment Information/Brief 2025-26 | UOS

Assessment Information/Brief 2025-26

To be used for all types of assessment and provided to students at the start of the module. Information provided should be compatible with the detail contained in the approved module specification.

Assessment task details and instructions

General Guidance for Completing the Coursework:

This coursework is designed to give you the opportunity to apply the machine learning and data mining techniques covered in this module. You will analyze datasets of your choice, draw conclusions from your analysis, and present your results in the form of a structured report.

The assignment consists of three tasks. For each task, you must select an appropriate dataset that fits the type of data mining problem specified. For example, for Task 1 (classification), choose a dataset that includes a target variable suitable for classification. You must use a different dataset for each task, and none of your chosen datasets should be the ones used in the workshops.

For each task, you are required to fully explain and document your experimental process, including: Exploratory Data Analysis (EDA), data preparation and cleaning, choice of algorithms and parameters, discussion and interpretation of the results.

You should complete a report for each of the three tasks and combine them into a single PDF document for submission. The total submission should be approximately 6,000 words (excluding references).
A report framework is available on Blackboard to guide you on the expected structure.

All code must be clearly commented, and screenshots of both code and outputs should be included in your report with appropriate explanations. Ensure that the images are clear and readable, as you will not receive marks for that section if the required screenshots are not included

The detailed requirements for each task are outlined as follows

Task 1 (40 Marks):

Choose a dataset suitable for a classification problem that includes at least 1,000 samples and 7 features. It can be any real-world or benchmark dataset that you have permission to access and use. For finding datasets, you can refer to the list of suggested repositories on page 6 of this document (You can use other resources as well).

You are required to complete the following steps and provide explanations for each in your report:

• Describe your dataset and provide appropriate references, perform exploratory data analysis (EDA), and preprocess the data as needed to prepare it for modeling. Additionally, discuss any ethical or social issues related to your dataset.
• Apply two classification algorithms of your choice on your chosen dataset using Python. Provide detailed explanations of your data mining processes, including the rationale behind the choice of algorithms and parameters.
• Present the results, compare the performance of the two algorithms, discuss and justify performance metrics.
• You should critically evaluate the classification output and discuss how it will benefit the related business or help solve the problem at hand. Explain how the insights gained from the model could influence decision-making, or improve processes.
• Suggest ways the business or organization can leverage these insights to achieve specific goals or improve performance (Provide actionable recommendations)
• Your report for this section should be structured and numbered according to the writing framework provided on Blackboard.
• You should include the screenshots of all the codes and their outputs/results with explanation in your report. Ensure that the images are clear and readable, as you will not receive marks for that section if the required screenshots are not included

Task 2 (35 Marks):

Choose a dataset suitable for a clustering problem that includes at least 1,000 samples and 7 features. It can be any real-world or benchmark dataset that you have permission to access and use. For finding datasets, you can refer to the list of suggested repositories on page 6 of this document (You can use other resources as well):

You are required to complete the following steps and provide explanations for each in your report:

• Describe your dataset and provide appropriate references, perform exploratory data analysis (EDA), and preprocess the data as needed to prepare it for modeling. Additionally, discuss any ethical or social issues related to your dataset.
• Apply two clustering algorithms on the selected dataset of your choice using Python. Provide detailed explanations of your data mining processes, including the rationale behind the choice of algorithms and parameters.
• Present the results, compare the performance of the two algorithms, and critically evaluate the clustering output
• Discuss how it will benefit the related business or help solve the problem at hand. Explain how the insights gained from the model could influence decision-making, or improve processes.
• Suggest ways the business or organization can leverage these insights to achieve specific goals or improve performance (Provide actionable recommendations)
• Your report for this section should be structured and numbered according to the writing framework provided on Blackboard.
• You should include the screenshots of all the codes and their outputs/results with explanation in your report. Ensure that the images are clear and readable, as you will not receive marks for that section if the required screenshots are not included

Task 3 (25 Marks):

Choose a dataset suitable for a sentiment analysis problem that includes at least 1,000 samples. It can be any real-world or benchmark dataset that you have permission to access and use. For finding datasets, you can refer to the list of suggested repositories on page 6 of this document (You can use other resources as well):
You are required to complete the following steps and provide explanations for each in your report:

Describe your dataset and provide appropriate references, perform exploratory data analysis (EDA) including showing wordcloud and preprocess the data as needed to prepare it for modelling, discuss ethical/social issues related to the dataset.

• Apply sentiment analysis on the selected text dataset of your choice using Python and provide detailed explanations of your approach.
• Present and analyze your results and discuss how it will benefit the related business or help solve the problem at hand. Explain how the insights gained from the model could influence decision-making, or improve processes.
• Suggest ways the business or organization can leverage these insights to achieve specific goals or improve performance (Provide actionable recommendations)
• Your report for this section should be structured and numbered according to the writing framework provided on Blackboard.
• You should include the screenshots of all the codes and their outputs/results with explanation in your report. Ensure that the images are clear and readable, as you will not receive marks for that section if the required screenshots are not included

Tips for Effective Dataset Selection:

For each task, you are free to select any dataset that meets the specified requirements mentioned in each task. You may use any public dataset or a dataset for which you have permission, ensuring that you consider any ethical or legal aspects

While technical suitability (e.g., minimum sample size and features) is important, you are also encouraged to choose data that aligns with your academic interests, career goals, or personal curiosities. Selecting a dataset that genuinely interests you can make the analysis more engaging, help you connect your technical work to real-world problems, and allow you to showcase your strengths in areas you are passionate about.

Please note that you must use a different dataset for each task, and none of your chosen datasets should be the ones used in the workshops.

To help you get started, some repositories that provide public datasets suitable for data mining tasks are listed below. You are welcome to use them if you wish or you can use any other resource. You may want to choose ones in domains you have existing experience or interest in.

We recommend avoiding very large datasets.

Using Generative Artificial Intelligence (GenAI) tools

You may use AI tools to support you in revising your report. However, if you do, you should include the following declaration at the beginning of your submission, so we know how these tools have been used:

During the preparation and writing of this submission I used [NAME TOOL / SERVICE] in Section [SECTION(s) TITLE]

in order to [REASON]. After using this tool/service, I reviewed and edited the content as needed and take full responsibility for the content of the submission content.

You are also reminded that you will need to provide citations if you reference published research, and AI tools cannot be relied on to provide accurate citations.

We would also emphasise that poor use of AI tools (e.g., copying and pasting output from ChatGPT that you don’t understand and without editing) can be clearly identified and will result in a low-quality submission which will score poorly.

Word count/ duration (if applicable)

Your assessment should be 6000 words (+/- 10%) in total across the three tasks.

How to submit

You  are  required  to  submit  your  coursework  in  the Assessment folder on Blackboard. You need to upload your written report as a single PDF file named "Surname_studentid.pdf" to the Submission area.

You also need to create and upload a zip file named "Surname_studentid.zip" containing your Python code in either .py or .ipynb format. The notebook should contain clear and well-structured comments for each section.

Do not upload links in the submission area, as we cannot mark your work. If you have saved your files on your University of Salford OneDrive, dragging files to the submission area may attach a link instead of the file. Always attach files using the paperclip icon.

Students with a Reasonable Adjustment Plan (RAP) or Carer Support Plan should check the plan to see if an extension to this submission date has been agreed.

Feedback

Feedback will be available on Blackboard within 15 working days after the assignment deadline. To support you with your assessment, drop-in sessions will be held on Fridays from 1–2 pm in room 2.18 from week3, where one of the lecturers will be available to answer your questions about the assignment.

Assessment criteria

You should look at the assessment criteria to find out what we are specifically looking at during the assessment. The assessment criteria for this assignment are detailed in the assignment rubric available on Blackboard.

Assessed intended learning outcomes

On successful completion of this assessment, you will be able to:

1. Design and implement data mining solutions to address practical problems, selecting and justifying appropriate methodologies.
2. Demonstrate knowledge of fundamental data mining methods such as classification, clustering, and text mining, and apply these techniques to real-world datasets.
3. Critically assess the performance and validity of data mining models, and interpret the results to make data-driven decisions.
4. Demonstrate proficiency in using data mining tools for data preprocessing, model development, and evaluation.
5. Present data mining results and insights clearly and effectively through written reports.

Employability skills developed / demonstrated

You will develop a range of employability skills sought by employers through each assessment.

Through this assessment will have an opportunity to develop and demonstrate the following employability skills:

I = You will have been introduced to this skill

U = You will have developed an understanding of this skill in the context of your subject

A = You will be able to apply this skill in the context of your subject

D = You will have demonstrated an enhanced understanding and application of this skill in a wider context

Reassessment arrangements

If you fail your assessment, and are eligible for reassessment, you will be able to find the date for resubmission on your module site in Blackboard. There is no resubmission if you are on a retake attempt.

For students with accepted personal mitigating circumstances for absence/non submission, this will be your replacement assessment attempt.

Your reassessment task will be the same as this assessment brief.

We know that having to undergo a reassessment can be challenging however support is available. Have a look at all the sources of support outlined earlier in this brief.