W1 Setup · Tooling & the big picture
Focus
Set up the tools you'll use throughout the course: a Python environment (Jupyter, Colab, or VS Code), an optional SAS Viya for Learners account if you want to try the enterprise track, and an AI assistant (Claude, ChatGPT, or Copilot) for pair-programming.
Orientation pages (skim, don't memorize)
- Analytics Big Picture — the whole discipline on one page.
- Analytics Use Cases Across Industries — where analytics actually lives.
- Four Types of Analytics — descriptive, diagnostic, predictive, prescriptive.
- Analytics Terminology Guide — vocabulary you'll hear in interviews.
- Different Decisions Need Different Tools — match decision to technique.
W2 "What's the real problem?" · Data Preparation & EDA
Decision question: Is this even an analytics problem?
Topic page
Focus
Before any modeling, you need to translate a business problem into a data problem and explore the data thoroughly enough to know what's there.
W3 EDA Workshop · Apply It to Fresh Data
Topic page
Focus
Apply EDA to a fresh dataset end-to-end — profile, clean, and write up three things you'd investigate before modeling.
Due
- Practice exercise: EDA
- Optional: IBM SkillsBuild — "What is Data Science?"
W4 "How much?" · Linear & Multiple Regression
Decision question: What should we price? What will they spend?
Topic page
Focus
When the thing you want to predict is a number, regression is your starting point. We cover simple and multiple linear regression and the assumptions you have to check.
W5 "Keep or let go?" · Logistic Regression & Classification
Decision question: Which customers will leave?
Topic page
Focus
The customer-churn classic. We use logistic regression to predict yes/no outcomes and learn how to read confusion matrices and ROC curves.
W6 "Keep or let go?" · Decision Trees & Interpretation
Decision question: How do we explain the model to executives?
Topic page
Focus
Decision trees produce rules anyone can follow. Great for explainability and for understanding feature interactions. Stent Failure case study.
Due
Practice exercise: Supervised methods
W7 "Keep or let go?" · Random Forests & Ensembles
Decision question: Can we improve accuracy?
Topic page
Focus
Single trees overfit; many trees together (random forests) usually win. We discuss bias–variance, ensembles, and when accuracy gains are worth the loss in interpretability.
Due
Optional: IBM SkillsBuild — Data Science with Python
W9 Supervised Workshop + Exam Review
Decision question: Can we trust this? Is the model good enough?
Topic page
Focus
Bring everything together: train/test splits, cross-validation, comparing models, and judging whether a model is "good enough" for the business decision.
W10 "Who are they?" · K-Means & Hierarchical Clustering
Decision question: What natural groups exist in our customers?
Topic page
Focus
Switch from supervised (we tell the model the answer) to unsupervised (the model finds patterns on its own). K-means and hierarchical clustering, applied to bank customer segmentation.
W11 "What's unusual?" · Anomaly Detection & Market Basket
Decision question: What doesn't fit? Is this fraud?
Topic page
Focus
Find the needles, not the haystack. Anomaly detection on healthcare vitals, sports performance, and streaming data. Market basket analysis on transaction data.
Due
Practice exercise: Sup + Unsup
W13 Self-Check · Concepts & Methods (optional)
Due
Self-check: concepts & methods
W14 "What are they saying?" · Text Analytics & Sentiment
Decision question: What's the sentiment in customer reviews?
Topic page
Focus
Most business data is text — reviews, support tickets, surveys, contracts. We tokenize, vectorize, score sentiment, and pull customer-review insights using both no-API and API workflows.
Due
Practice exercise: Text analytics
W15 "What's next?" · Time Series Forecasting
Decision question: What will sales be next quarter?
Topic page
Focus
Trend, seasonality, and forecasting. ATM cash-demand forecasting, equipment monitoring, and the FreshMart demand-planning case.
Due
- Optional: IBM SkillsBuild — Data Analysis
- Optional: IBM SkillsBuild — Data Visualization
W16 Capstone · End-to-End Integration
Decision question: Prove you can do this.
Cases page
Focus
If you've worked through one of the integrative cases or built your own end-to-end project, this is your moment to share it. The deliverable is a working artifact plus a short walkthrough.
Due
Capstone project
W17 FINALS — Take-home Exam (applied case)
Take-home self-check: an applied SAS-based case using the techniques you've learned throughout the course.
Due
Self-check: applied case