Hello, if you have any need, please feel free to consult us, this is my wechat: wx91due
SEMESTER 2 2024/25
COURSEWORK BRIEF:
|
Module Code: |
MANG6554 |
Assessment: |
Individual Coursework |
Weighting: |
100% |
||||||||||
|
Module Title: |
Advanced Analytics |
||||||||||||||
|
Submission Due Date: @ 16:00 |
23rd May 2025 |
Word Count: |
2500 |
||||||||||||
|
Method of Submission: |
Electronic via Blackboard Turnitin ONLY (Please ensure that your name does not appear on any part of your work) |
||||||||||||||
|
Any submitted after 16:00 on the deadline date will be subject to the standard University late penalties (see below), unless an extension has been granted, in writing by the Senior Tutor, in advance of the deadline. Days Late: Mark: 1 (final agreed mark) * 0.9 2 (final agreed mark) * 0.8 3 (final agreed mark) * 0.7 4 (final agreed mark) * 0.6 5 (final agreed mark) * 0.5 More than 5 0 |
|||||||||||||||
|
This assessment relates to the following module learning outcomes:
|
|||||||||||||||
Coursework Brief:
(30/100 marks) Part A1.
As an analyst, you will explore sentiment analysis techniques for a company that sells city passes. These passes provide customers access to multiple vacation spots, such as museums, zoos, and cruise rides. The company collects customer feedback on each location and stores it in separate datasets. The datasets are labelled, with 1 indicating positive sentiment and 0 indicating negative sentiment.
Your task is to implement at least four different models, including BERT, and compare their performance. Justify your choice of methods, explain your evaluation approach, and determine which model(s) perform best.
(30/100 marks) Part A2.
Users of the city pass app can send invitations to other users, such as friends, to encourage them to travel together.
Your task is to build a predictive model using graph neural networks (GNNs) to predict the success of these invitations. A summary of the data files and variables is provided in the table below.
In your solution, explain how the graph neural network is designed, the reasons for adopting a specific architecture, and the model's performance. Additionally, discuss how the predictive model could be used to promote more travel opportunities for users and propose methods to test the effectiveness of these strategies.
|
Data file |
Variable name |
Description |
|
User data |
uid |
Unique Identifier of the Users: A distinct ID assigned to each user. |
|
User data |
user_profile |
User Information: Details from the user’s profile page, typically including information about their travel experiences or preferences. |
|
Message data |
sid |
Sender User ID: The unique identifier of the user sending the message. |
|
rid |
Recipient User ID: The unique identifier of the user receiving the message. |
|
|
message |
Message Text: The content of the invitation sent by the sender to the recipient. |
|
|
success |
Outcome (1 or 0): A binary variable indicating the result of the invitation: 1 if the sender and recipient travel together using the city pass, or 0 if they do not. |
(30/100 marks) Part B.
Please use ten years of data from 10 assets selected from NASDAQ shares (preferably using a systematic asset selection approach) within the date range 2015-01-01 to 2025-01-01. Apply the necessary data pre-processing and processing techniques to:
• Develop a momentum strategy as a basic algorithmic trading strategy.
• Backtest the developed trading strategy and evaluate the pros and cons of your strategy.
• Optimize your strategy by incorporating stop-loss and take-profit mechanisms and evaluate its performance and robustness.
• Compare the results across different shares and industries and identify which one recovered from COVID-19 in the shortest time.
• Construct a portfolio and optimise it using different scenarios.
(10/100 marks) Well-structured writing with accurate use of language
Structure, formatting, reporting style and concise presentation are expected for the report.
Helpful tips:
You may contact [email protected] for part A and [email protected] for part B.
For all Parts:
You are recommended to use the writing style suggested below and please be consistent about these settings.
• Fixed font and size: Times new roman, size = 11 for body text
• Use headings for different sections
• 1.5 line/paragraph spacing
• Add captions to the tables and figures
• Use good-quality images for figures
• Cite references properly using Harvard
• For reproducibility, please set random seed using your student number
• The Programming language: Python, others are not allowed.
Carefully report the various steps of your methodology and discuss your results in a rigorous way! Please do not include code in your report. You can put your code (in ipynb or py format) into a zip file and submit via Turnitin using an additional file submission link, other than the “REPORT ONLY” link.