Machine Learning for Analysts: Models & Evaluation is a structured track of 13 lessons that build a complete, interview-ready understanding of the topic. Work through them in order, then use the quiz and flashcards in each lesson to revise.
What this course covers
- What Machine Learning Is, and When a Query Is Enough - Where a rule, an average or a query beats a model - the judgement of a senior analyst.
- Supervised versus Unsupervised Learning, With Examples - The two families, and which business problems fall into each.
- Train, Validation & Test Splits, and Cross-Validation - Why three splits, how leakage happens, and when cross-validation is worth it.
- Overfitting, Underfitting & the Bias-Variance Trade-off - Diagnosing which problem you have from the training and validation curves.
- Feature Engineering: Where Most of the Gain Comes From - Encoding, scaling, interactions and time-based features, with a worked improvement.
- Linear Regression: Fitting, Interpreting & Diagnosing - Reading coefficients correctly, checking assumptions and spotting a bad fit.
- Logistic Regression & Interpreting Probabilities - Odds, log-odds and turning a predicted probability into a business decision.
- Decision Trees, Random Forests & Gradient Boosting - How each works, why boosting usually wins on tabular data, and the cost of that.
- Clustering & Segmentation: Choosing the Number of Clusters - Distance-based clustering, picking the cluster count, and making segments usable.
- Classification Metrics: Precision, Recall & the Confusion Matrix - Every cell of the confusion matrix, and choosing the metric that matches the cost.
- Regression & Ranking Metrics, and Picking the Right One - Error metrics compared, and why the default is often the wrong choice.
- Explaining a Model to a Business Audience - Feature importance, worked examples and analogies that survive a leadership meeting.
- Case Study: Building & Evaluating a Churn Model - A churn model built, evaluated and translated into a retention action.