辅导 DATA4800 Artificial Intelligence and Machine Learning Assessment 3讲解 迭代

Assessment 3 Information

Your Task

Develop a real-world Machine Learning or AI project plan/proposal based on the learnings from the subject.

Assessment Description

This assessment seeks to simulate a real-world task that you may have to undertake in the future. Therefore, the assignment is non-prescriptive and requires you to pose a relevant, small, creative and significant problem to solve that could result in benefits to the organisation of choice.

In this assessment, you need to consider an organisation in an industry of your choice and articulate the steps this organisation needs to take to enable Machine Learning and/or AI for data-driven decision making. You are required to analyse a sample data set to demonstrate  expected AI/ML outcomes.

You need to be familiar with the organisation and industry (e.g., where you have worked or are  working, a future start-up company), NOT an organisation such as Amazon/Boeing/Qantas etc.

Well-reasoned use of Generative AI is encouraged. However, generic and irrelevant content will be heavily penalised in the marking.

The report should address:

o Why AI would help this organisation given their current operations

o What Machine Learning techniques you would recommend

o An example of the predictive model using sample data

o Deployment considerations for the model

o The benefits for the organisation clearly articulated with estimates of expected revenue/profits or Return on Investment

Assessment Instructions

You will be asked to produce a report and video for this assessment.

PART A: Report (25 marks)

•    By Week 9 identify a company and industry you are familiar with that would benefit from Machine Learning/AI. Define a business problem that can be solved using

Supervised Machine Learning - Classification in the chosen company (binary or multi- class).  Find a sample dataset suitable to solve the business problem defined. Note:

o The application needs to be based on Machine Learning/AI (not some other   aspect of analytics). Do not select a regression, forecasting, or reinforcement learning task.

o Focus on a single, well defined (small) application.

o Sample datasets maybe sourced from:

   an organisation. if you work there

   public repositories

   Open government data

•   The company, business problem, and dataset must be validated by your facilitator before you proceed with other steps.

•    By Week 12 draft some preliminary points about the report in class. You are

encouraged to consider the current mode of operation, possible inefficiencies,

available data and how this data may be used to provide efficiencies based on the

concepts and techniques covered in the subject. Think of yourself as a consultant or a founder.

•    Include a list of references that are directly related to the content. Each reference

needs to be linked to at least one specific point in the content of your assessment. It is expected that you will have at least six relevant references.

•    Upload the files that contain your predictive modelling workflow (in Orange) to the file   submission Dropbox provided on the assessment page. No marks will be awarded for  the assessment unless the report and the Orange workflow files have been submitted.

•   The report must be written using Google docs template (shared by your lecturer) and submitted via Turnitin. To properly use the Google Docs template, please follow the   below steps:

1. Go to https://docs.google.com/document/d/1enPWUYRaZj-4nRIYbPX7ZvPBQaoKUDQFkmmuZWTCWFU/edit?usp=sharing

2. Sign in to your Gmail/Google account (if not already signed in). Click “File – Make a Copy” to copy the template to your Google Drive

3. Click “Share” . Change “Restricted” to "Anyone with the link”, and change “Viewer” to “Editor” .  Click “Copy link” .

4. Paste the copied link in the header of your Google Doc under “Kaplan Business School” .

5. Do all your writing directly in your Google Doc under your Gmail/Google account.

6. Download your report as PDF and submit via Turnitin by the submission deadline.

PART B: Video Quiz (15 marks)

•    Record yourself doing the video quiz using zoom screen share. Instructions are available on the assessment page.

•     There are four sections: (1) business problem identification, (2) data collection, (3)

machine learning implementation, and (4) improvements. For each section, the student will have 2-3 minutes to answer all the questions within the section. The total allocated  time for the video quiz is 15 minutes.