EFIM30051 Data Analytics and Artificial Intelligence for Business
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
EFIM30051
Data Analytics and
Artificial Intelligence for Business
Assignment Brief
Assignment Part 1 – 1,000 words (50%)
This part of the assignment will assess you against the following Intended Learning Outcomes (ILOs):
ILO 1: demonstrate a systematic and critical understanding of the areas of application of predictive models in organisations and their use in Artificial Intelligence;
ILO 2: analyse a business problem and identify and apply appropriate predictive analytic techniques to provide solutions.
Task
Identify and select secondary data sources on a topic which interest you. These may be sources available at the library, sources available from government sites or other open sources available through internet searches.
You may also use data sources which appear in the tutorial sections of Blackboard. It is important NOT to sue the same data and analysis method which has been covered by the tutor in the lectures or tutorials. I.e., you need to show you have done your own work and not handed in an assignment which is similar to one demonstrated in a lecture! Also, some data sets used in class are simplistic in order to demonstrate a specific method so please do not use these. These data sets will NOT appear in the tutorial sections of Blackboard, instead in the lecture sections. If you have questions, please contact Andrew Rogers via email to check.
Identify/explore the data and determine how it can be used to derive relevant insight/insights around a specific issue you think is relevant (ILO 1). Use some of the techniques from this course to analyse and apply/build appropriate analysis/models which would help address the issue in question (ILO 2).
Discuss these analyses/models in terms of how they perform in terms of statistical robustness AND how the analyses/models can be used to convey to a lay audience (e.g., stakeholders, management, etc.) any findings/insights from the work. Highlight any limitations with the data/analysis/models and implications/consequences to the lay audience (ILO 2).
During the course you learned how statistical models can be used to automate business processes and decision making through machine learning. For example, logistic regression is a classification algorithm used in credit scoring to decide whether an applicant qualifies for a loan application or not. Discuss how suitable the data/model would be to be adapted to a machine learning algorithm to automate a business process or decision making (ILO 1).
Data sources
A Google search of “free secondary data sources” OR “regression data sets” etc. are just two examples which returns a wide range of sources (some of which are below though do not be limited to just these). You may choose a search more specific to your interest.
It is important you spend time in researching appropriate data sets which match your interests or even which can be used to better inform you with regards to future academic research. Selecting the data is probably just as important to the assignment as the analysis and so you should be encouraged to devote some time to this element of the assignment. Remember you may use some of the course data sets (see note earlier).
https://www.ons.gov.uk/ (Office of National Statistics (UIK))
https://jbh.co.uk/blog/content-creation/100-free-data-sources-digital-pr-campaigns/ (list of many data sources and links)
https://guides.lib.vt.edu/c.php?g=580714 (Data resources for social science: Research datasets for secondary analysis)
https://nsufl.libguides.com/c.php?g=111860&p=4691028 (Public Health: Secondary Data Analysis/Research: Statistics and Data Science)
http://www.secondarydata.com/
https://www.dataquest.io/blog/free-datasets-for-projects/
UCI Machine Learning Repository: Data Sets
Find Open Datasets and Machine Learning Projects | Kaggle
(Kaggle requires you to register for free. I used my Bristol email address)
Useful databases can also be found from the Library website. You just need to filter by subject: http://bristol.ac.uk/library/find/databases
Assignment Part 2 – 2,000 words (50%)
This part of the assignment will assess you against the following ILOs.
ILO 3: critically evaluate the impact of AI on businesses and in the workplace;
ILO 4: provide a reasoned analysis and evaluation of the main ethical and societal implications of AI;
ILO 5: critically evaluate a case for the adoption of AI to solve a business problem.
Task
Choose ANY ONE of the following business problems and write a 2,000-word critical evaluation of your proposed AI solution. Your evaluation should:
a. address the requirements of the company who intends to adopt the AI solution;
b. draw on theories you have learned in the unit and independent research.
Business problem 1 – Digital fitness platform
ULTRAFit is a chain of 105 gyms in the UK, employs about 300 trainers, and counts more than 200,000 registered members. During the COVID-19 pandemic, ULTRAFit had to move its operations online and provide digital fitness sessions. This proved to be a success as demonstrated by a steady growth of its membership. Following this positive experience, ULTRAFit is looking into extending the digital fitness area of its business through the adoption of a new digital platform for the delivery of training sessions online and on-demand. You have been tasked to write a report on how AI/ML can be integrated into the platform to (i) maximise value; (ii) increase customer engagement, and (iii.) deliver a personalised digital fitness experience. Your report should cover the following:
1. how your proposed AI/ML solution(s) help(s) ULTRAFit achieve (i); (ii); and (iii) above;
2. the impact of the proposed AI/ML solution(s) on ULTRAFit’s workforce (e.g. in terms of size, skills, morale);
3. potential ethical concerns that ULTRAFit’s customers might have and how to address them;
4. a conclusion section where you develop a balanced case for (or against) adopting AI/ML in ULTRAFit’s digital platform based on your evaluation. If there is a case for adoption, summarise key recommendations to maximise benefits and minimize risk.
Business problem 2 – Digital Wellbeing in the Workplace
You are the HR manager of the Employees Wellbeing Programme of an Accountancy company in the UK. As it often happens with most clerical jobs, employees of this company spend most of their time at work sat at a desk and in front of a computer. The CEO of the company has read reports about the negative impact of sedentary behaviour in the workplace including increased risk of diabetes, obesity, and mental health issues. On the long run, poor mental and physical wellbeing of employees could have negative consequences for workplace productivity and performance. The CEO has asked you to write a report on a digital wellbeing intervention that uses AI/ML to reduce sitting behaviour and improve employees’ physical and mental wellbeing. Your report should cover the following:
1. how your proposed AI/ML solution helps employees reduce sedentary behaviour and improve their mental and physical wellbeing;
2. the impact of the proposed AI/ML solution on the work, performance, and productivity of employees and how to ensure that employees fully accept and engage with the digital wellbeing programme;
3. potential adverse ethical effects on employees resulting from the adoption of the proposed AI/ML solutions and how these can be mitigated;
4. a conclusion section where you argue for a balanced case for (or against) adopting AI/ML to improve wellbeing in the workplace. If there is a case for adoption, summarise key recommendations to maximise benefits and minimize risk.
Business problem 3 – Insurance sales automation
CLH Insurance is an international insurance group with around 7,000 sales agents and over 25 million customers across France, Germany, Hong Kong, Japan, and the UK. CLH Insurance is looking into adopting AI/ML to automate and streamline its insurance sales operations. You have been tasked to write a report on how AI/ML can help the company achieve the following: (i) improve the efficiency of its sales operations (e.g. in terms of customer queries processed) and decrease pressure on its sales agents; (ii) increase the effectiveness of its sales team (e.g. in terms of number of sales and customer satisfaction); (iii) improve the personalisation of products and quality of services.
Your report should cover the following:
1. how your proposed AI/ML solution(s) help(s) CLH Insurance achieve (i); (ii); and (iii) above;
2. the impact of the proposed AI/ML solution(s) on CLH Insurance’s workforce (e.g. in terms of size, skills, morale) and how to ensure wide acceptance and effective use of the AI/ML solution(s) by employees;
3. potential adverse ethical effects on employees and/or customers resulting from the adoption of the proposed AI/ML solution(s) and how these can be mitigated;
4. a conclusion section where you develop a balanced case for (or against) adopting AI/ML for the automation of CLH Insurance’s sales operation based on your evaluation. If there is a case for adoption, summarise key recommendations to maximise benefits and minimize risk.
Structure and formatting
Your report will have to be written in an academic style. This means that you are expected to use theories learned in the course and independent reading in the analysis. Your report will have an introduction, main body, and conclusion. You are free to use any font and text formatting provided that it is clear to read.
Referencing
You will have to cite and reference your sources following accepted academic standards. Please find further guidance on the Library’s website.
DEADLINE: h 13:00 (UK time) on Monday 13th December 2021
Marking criteria
Part 1:
1. The extent you have identified the issue, identified and acquired appropriate data to help address the issue in hand. Clarify on the proposed outcome and how this benefits the research area.
2. The extent to which you have selected the analysis method(s) and the appropriateness of any models you have selected. The extent to which the analysis techniques/models are applied to the data and how the outputs are interpreted.
Part 2:
3. The extent to which the proposed AI/ML solution addresses the company’s requirements.
4. The extent to which theories learned during the lectures and through independent research have been properly selected and applied.
5. The level of critical evaluation of the potential impact of the proposed AI/ML solution in relation to the business area of application, workforce, and ethics.
6. The extent to which the argument in support of (or against) the adoption of AI/ML that is logical, convincing and well supported.
First (70+) |
2:1 (60-69) |
2:2 (50-59) |
Third (40-49) |
Marginal fail (35-39) |
Outright fail (0-34) |
|
Knowledge and understanding |
Part 1: Excellent comprehension of the issue identified and ability to identify and acquire appropriate data to help address the issue at hand. Excellent clarity in explaining the proposed outcome and how this benefits the research area. |
Part 1: Very good comprehension of the issue identified and ability to identify and acquire appropriate data to help address the issue at hand. Very good clarity in explaining the proposed outcome though not always displaying an understanding of how this benefits the research area. |
Part 1: Generally clear and accurate comprehension of the issue identified, through there may be some errors and/or gaps in understanding. Data have been properly identified and acquired but have some limitations in addressing the issue at hand. Clarity in explaining the proposed outcome, though with limited understanding of how this benefits the research area. |
Part 1: Limited comprehension of the issue identified, with significant errors and omissions. Data have flaws in the way they have been identified and acquired and do not properly address the issue at hand. Generally ignorant or confused explanation of the proposed outcome and how this benefits the research area. |
Part 1: Unsatisfactory comprehension of the issue identified. Data are inadequate and fail to address the issue at hand. Limited or no explanation of the proposed outcome and how this benefits the research area. |
Part 1: Very limited, and seriously flawed comprehension of the issue identified. Insufficient or no data have been acquired. Limited or no outcome is proposed. |
Analysis |
Part 1: Selected analysis methods and models are highly appropriate. Excellent ability to apply analysis techniques/models to data and interpretation of outputs. |
Part 1: Selected analysis methods and models are appropriate. Very good ability to apply analysis techniques/models to data and interpretation of outputs. |
Part 1: Selected analysis methods and models are not entirely appropriate and have some limitations. Good ability to apply analysis techniques/models to data, though the interpretation of outputs is not always accurate. Structure may not be entirely clear or logical. |
Part 1: Selected analysis methods and models have severe limitations. Poor ability to apply analysis techniques/models to data. Interpretation of outputs is mostly inaccurate. Underdeveloped or chaotic structure. |
Part 1: Selected analysis methods and models are inappropriate. Little or no ability to apply analysis techniques/models to data. Limited or no interpretation of outputs. Lacking a coherent structure. |
Part 1: Selected analysis methods and models are inappropriate and flawed. No ability to apply analysis techniques/models to data and no interpretation of outputs. No attempt to provide a structure. |
Sources |
Part 2: Evidence of reading widely beyond the prescribed reading list and creative use of theories to enhance the overall argument. |
Part 2: Clear and generally critical knowledge of relevant literature; use of works beyond the prescribed reading list; demonstrating the ability to be selective in the range of material used, and the capacity to synthesise rather than describe. |
Part 2: Good attempt to go beyond or criticise the ‘essential reading’ for the unit; but displaying limited capacity to discern between relevant and non-relevant material. |
Part 2: Limited, uncritical and generally confused account of a narrow range of sources. |
Part 2: Little use of sources and what is used reflects a very narrow range or are irrelevant and/or misunderstood. |
Part 2: Limited, uncritical and generally confused account of a very narrow range of sources. |
Writing style and presentation |
Extremely well presented: minimal grammatical or spelling errors; written in a fluent and engaging style; exemplary referencing and bibliographic formatting. |
Very well presented: no significant grammatical or spelling errors; written clearly and concisely; fairly consistent referencing and bibliographic formatting. |
Adequately presented: writing style conveys meaning but is sometimes awkward; some significant grammatical and spelling errors; inconsistent referencing but generally accurate bibliography. |
Poorly presented: not always easy to follow; frequent grammatical and spelling errors; limited attempt at providing references (e.g. only referencing direct quotations) and containing bibliographic omissions. |
Unsatisfactory presentation: difficult to follow; very limited attempt at providing references (e.g. only referencing direct quotations) and containing bibliographic omissions. |
Very poorly presented: lacking any coherence, significant problems with spelling and grammar, missing or no references and containing bibliographic omissions. |
2021-12-05