BUS5PA Predictive Analytics Part C

Part C: Case Study - Predictive Analytics Life Cycle (30%- Maximum 3 pages)

A national retail chain has experienced a significant decline in customer retention over the past year. Senior management aims to implement predictive analytics to identify potential churners, segment customers for targeted marketing campaigns, and forecast future sales trends. The business has collected customer transaction history, loyalty scores, demographics, and online engagement data; however, no detailed analysis has been conducted yet.

You have been asked to write a report outlining how predictive analytics techniques can be used to address these challenges, using the Predictive Analytics Life Cycle as your framework. Relate to the Predictive Analytics Life Cycle to answer the following questions:

1. Stage 1: Identify the problem

a) Define the business problem: Why is customer retention important for the business? b) Translate the problem into an analytics objective.

2. Stage 2&3: Data preparation/transformation

a) Hypothesize and write what sort of data sources should be collected. b) Describe what sort of data preparation tasks will be required. c) Discuss the importance of feature engineering at this stage.

3. Stage 4: Model development:

Discuss the following predictive analytics techniques and what particular purpose each of these techniques are used for in developing an analytics-based solution within the retail case: I. Decision Trees II. Logistic Regression III. Neural Networks IV. K-Nearest Neighbours (KNN) V. Clustering VI. Time series analysis

4. Stage 5: Model Evaluation and Business Recommendations

a) Explain how predictive models should be evaluated and justify the importance of model evaluation. b) Provide three practical recommendations the business could act on if these models were implemented successfully.