1 / 12

DATA4400 Assessment 3

Forecasting Techniques and Time Series

Individual Project (40% of total grade)

Presentation: 10% | Report: 30%

Due: Week 12 - Presentation in class + Report via Turnitin

Real-world forecasting project that simulates professional consulting work

What You'll Do

Part A - Presentation (10%)

  • Duration: 3-8 minutes
  • Content: Company overview, business problem, forecasting solution
  • Format: 5-6 slides maximum
  • Submit: 5 minutes before class

Part B - Report (30%)

  • Length: 1000 words (+/-10%)
  • Format: Microsoft Word document
  • Include: Data files (zipped)
  • Submit: Via Turnitin by Friday 11:55 PM
Important: This is a non-prescriptive assignment requiring creativity and original thinking!

Choose Your Plan

Plan A - Your Workplace

Best Option! Use data from your current/previous employer. Can be scaled/modified for confidentiality.

Examples: Care home client data, grocery store waste patterns

Plan B - Public Data

Source genuine time series data and create realistic scenarios.

Examples: EV sales vs charging stations, tourism forecasting, public health data

Plan C - Financial Data

Use share prices from Yahoo Finance with business context.

Examples: Bank stock analysis, employee share schemes, Granger causality

Do NOT use: AI-generated datasets (mostly fictitious and not current) - Kaggle is accepted with a good source

Presentation Structure

Slides 1-2

Company & Problem

Industry context & forecasting challenge

Slides 3-4

Forecasting Solution

How forecasting helps & expected benefits

Slides 5-6

Data Overview

Dataset, variables, source relevance

Pro Tips: Keep visuals clear, focus on practical implications, engage your audience!

Report Structure (1000 words)

1. Organization Overview

Industry, challenges, decision-making processes

2. Importance of Forecasting

Value for decision-making, efficiency, performance

3. Time Series Data

Dataset, source, variables, descriptive analysis

4. Recommended Technique

Justify forecasting method (VAR, Prophet, etc.)

5. Results & Analysis

Visual results, error metrics, commentary

6. Business Benefits

Financial ROI and/or societal advantages

Project Timeline

Week 10

Plan Selection

Choose company & industry

Week 11

Draft Outline

Key points & methodology review

Week 12

Final Submissions

Present + Submit report

Week 12 Deadlines:
  • Presentation slides: 5 minutes before class
  • Report + data files: Friday 11:55 PM AEST

How to Prepare - Step by Step

Data Preparation

  • Source your time series data
  • Clean and format the data
  • Perform descriptive analysis
  • Identify trends, seasonality, anomalies
  • Create visualizations

Analysis & Modeling

  • Choose appropriate forecasting technique
  • Justify your method selection
  • Apply the forecasting model
  • Evaluate accuracy and interpret results
  • Connect findings to business value
Remember: You don't need to compare multiple methods or split data into train/test segments. Focus on meaningful interpretation!

Expected Outcomes

Technical Skills

  • Apply forecasting techniques to real data
  • Interpret forecast accuracy metrics
  • Create meaningful data visualizations
  • Use appropriate software tools (Excel, R, Python)

Professional Skills

  • Present complex analysis clearly
  • Write professional technical reports
  • Connect analysis to business value
  • Think like a consultant/analyst

Success Criteria

Your project should demonstrate how forecasting can drive real business improvements, whether financial efficiency, operational optimization, or strategic decision-making.

Where to Find Data

Workplace Sources

  • Annual reports (8+ years ideal)
  • Sales/demand records
  • Operational metrics
  • Customer data

Public Repositories

  • Government open data portals
  • GovHack datasets
  • Industry association reports
  • Economic indicators

Financial Data

  • Yahoo Finance (share prices)
  • Commodity prices
  • Economic indicators
  • Trading volumes
Data Quality: Ensure your data is current, genuine, and relevant to your business problem!

Generative AI Guidelines

AI Usage is OPTIONAL (Level 2)

You may use AI for research and content generation, but it must be properly referenced.

Requirements if Using AI:

  • Reference AI collaboration like any other source
  • Include appendix with all prompts and responses
  • Demonstrate original thinking
  • Avoid over-reliance on AI-generated content

Assessment Impact:

  • Penalties for over-reliance on AI
  • Must show your own analysis and insights
  • AI should enhance, not replace your work
  • Academic integrity review if challenged

Common Mistakes to Avoid

Scope Issues

  • Choosing overly complex or broad problems
  • Focusing on general analytics instead of forecasting
  • Using outdated or irrelevant data

Technical Issues

  • Poor data quality or insufficient cleaning
  • Inappropriate forecasting method selection
  • Lack of business context or interpretation

Submission Issues

  • Wrong submission platforms
  • Missing or late data files
  • Inadequate AI usage documentation

Tips for Success

Project Focus

  • Start small: Well-focused example beats complex analysis
  • Be creative: YouTube views, fruit waste - any interesting application!
  • Show impact: Connect forecasting to real business benefits
  • Quality over quantity: Simple Excel analysis can be excellent

Presentation Tips

  • Tell a story: Problem → Solution → Impact
  • Use visuals: Charts speak louder than numbers
  • Practice timing: 3-8 minutes goes quickly!
  • Engage audience: Make it interesting and relevant

Final Advice

Approach this as a consultant would - identify a real problem, propose a data-driven solution, and demonstrate clear value. Your facilitator is there to help with feasibility and suitability questions!