Reference no: EM133950200
Assessment Title - Skills-Building: Predictive & Prescriptive Analytics for Business Decision Making
Task
Follow the steps within the Google Colab notebook to complete the analysis and interpret the results. Using the outputs from your notebook analysis, write a comprehensive business report in .docx format that presents your findings and recommendations to business leadership.
Background
Imagine that you are a Business Analyst in a leading SaaS technology company and that you have been tasked with developing a comprehensive customer retention and revenue optimisation strategy. Get expert-level assignment help in any subject.
This skills-building exercise is a workflow that has been designed to simulate this use case and consolidate your practical knowledge of Python, Google Colab, and advanced machine-learning approaches for business analytics, including predictive modeling, time series forecasting, and causal inference.
You will use the notebook outputs to write a professional business report with data-driven recommendations.
Assessment Instructions
Phase 1: Notebook Analysis
1A: Data Exploration & Preparation
Run the provided code to load and explore the customer dataset
Execute data preparation steps for machine learning models
Note key insights about customer behavior patterns from the data exploration
1B: Predictive Modeling
Execute the LightGBM model for customer churn prediction
Run the Temporal Fusion Transformer model for revenue forecasting
Record model performance metrics and accuracy results for your report
1C: Explainable AI Analysis
Generate SHAP waterfall plots, force plots, and summary plots
Execute feature importance analysis
Identify the top 5 factors influencing customer churn and document insights about customer behavior
1D: Causal Analysis
Run the EconML causal learner for retention campaign analysis
Execute Average Treatment Effect (ATE) and Conditional Average Treatment Effect (CATE) analysis
Capture results about campaign effectiveness and which customer segments benefit most
Phase 2: Business Report Writing (1,200 words maximum)
Using the outputs from your notebook analysis, write a professional business report in .docx format with the following structure:
Business Problem & Data Overview (250 words)
Context of customer retention challenges
Description of the dataset and key customer metrics
Initial insights from data exploration
Predictive Analytics Findings (400 words)
Churn Prediction Results: Present LightGBM model performance and key findings
Revenue Forecasting Insights: Summarise TFT model predictions and business implications
Use charts/tables from your notebook to support your narrative
Explainable AI Insights (200 words)
Feature Importance Analysis: Discuss the top 5 factors driving customer churn based on SHAP analysis
Customer Behavior Patterns: Explain what the SHAP plots reveal about different customer segments
Causal Analysis & Treatment Effects (150 words)
Campaign Effectiveness: Present ATE results and the overall impact of retention interventions
Segment-Specific Results: Discuss CATE findings and which customer groups respond best to treatments
Recommendations (250 words)
Connect technical findings to business understanding for recommendations
Based on these findings, what is the recommendation?
Business Communication & Report Structure
Write a professional report with a compelling narrative and executive-level presentation quality.
Integrate analytical insights with compelling visual storytelling and data-driven arguments.
Report Requirements:
Format: Professional business report in .docx
Length: Maximum 1,200 words (excluding charts/tables)
Charts/Tables: Include 3-5 key visualisations from your notebook analysis
Citations: Reference any external sources using the Harvard referencing style
Professional Language: Write for a business executive audience and ask for funding from the executive board to solve the business problem.