← All courses

Course · 12 modules · 87 lessons

Data Analytics III: Machine Learning & AI Engineering

A 12-week hands-on course that takes you from "I can analyse data in pandas" to "I can build, evaluate and honestly report a predictive model and a grounded AI assistant." The first half is machine learning done properly: framing a prediction problem, choosing a baseline, regression and error metrics, gradient descent written from scratch, scikit-learn pipelines that never leak, classification with thresholds chosen by the real cost of mistakes, cross-validation, and a reproducible ML project evaluated once on a held-out test set. The second half is AI engineering: how language models work, prompts with output contracts validated in code, evaluation sets that catch regressions, retrieval with source citations that refuses to invent answers, and a read-only tool-using assistant with timeouts and failure handling. You finish with an integrated capstone — a real decision problem taken from question to data, dashboard or report, and a justified model or grounded assistant — defended in a 10-minute presentation. Entry requirement: Python functions, pandas, charts and basic statistics (Data Analytics II: Python for Data Analytics, or equivalent). Expect 10–15 hours per week of lessons, labs and a weekly project.

You can preview 3 lessons free.

Enroll for full access to every lesson, practical and instructor review.

Enroll for full access

01Week 1 — Prediction Problems & Data

  1. Using AI in this course: the rules
    Free preview20m
  2. What machine learning is (and is not)
    Free preview50m
  3. Framing a prediction: decision, timing, cost
    50m
  4. Is this data fit for prediction?
    45m
  5. Leakage: when the answer sneaks into the question
    45m
  6. Lab: set up and inspect three datasets
    Practical90m
  7. Lab: compute baselines by hand and in code
    Practical75m
  8. Weekly project: a written experiment plan
    ProjectFree preview180m

02Week 2 — Regression & Error

  1. Linear prediction: weights, bias and a line of best fit
    50m
  2. Residuals, MAE, MSE and RMSE
    50m
  3. Reading residuals and beating a baseline
    50m
  4. Many features: coefficients and their limits
    50m
  5. Lab: loss by hand, then in code
    Practical75m
  6. Lab: housing regression with residual plots
    Practical90m
  7. Weekly project: regression experiment report
    Project180m

03Week 3 — Training & Optimisation

  1. Loss and gradients: how a model learns
    50m
  2. Gradient descent from scratch in NumPy
    55m
  3. Learning rate, batch size and epochs
    50m
  4. Underfitting, overfitting and validation loss
    50m
  5. Lab: one parameter at a time
    Practical90m
  6. Lab: provoke and diagnose overfitting
    Practical75m
  7. Weekly project: a reproducible training experiment log
    Project180m

04Week 4 — Data Preparation & Pipelines

  1. Split first: train, validation and test
    50m
  2. Missing values, scaling and encoding
    50m
  3. Pipelines and ColumnTransformer
    55m
  4. Features, duplicates, outliers and imbalance
    50m
  5. Lab: watch leakage invent a model
    Practical75m
  6. Lab: a full preprocessing pipeline, scaled vs unscaled
    Practical90m
  7. Weekly project: a leak-proof pipeline for your problem
    Project180m
  8. When AI suggests a feature that leaks
    30m

05Week 5 — Classification & Metrics

  1. Logistic regression: scores, sigmoid and thresholds
    50m
  2. Confusion matrix, accuracy, precision and recall
    55m
  3. ROC curves, AUC and precision-recall curves
    50m
  4. Choosing a threshold by the cost of mistakes
    50m
  5. Lab: a breast-cancer classifier with deliberate thresholds
    Practical90m
  6. Lab: when 99% accuracy means nothing
    Practical75m
  7. Weekly project: classification report
    Project180m

06Week 6 — Experiments & Model Selection

  1. Cross-validation and hyperparameters
    50m
  2. Searching hyperparameters inside a pipeline
    50m
  3. Error analysis, subgroups and uncertainty
    50m
  4. Documenting experiments and avoiding causal claims
    45m
  5. Lab: two approaches, one split, honest selection
    Practical90m
  6. Lab: find where the model fails
    Practical75m
  7. Weekly project: experiment table and recommendation
    Project180m
  8. "The AI said my model is 99% accurate": checking claims
    25m

07Week 7 — Neural Networks, Embeddings & the ML Project

  1. Neural networks: layers of learned computations
    50m
  2. Reading training and validation curves
    45m
  3. Encodings and embeddings: sparse vs dense vectors
    50m
  4. Fine-tuning, distillation and scoping your ML project
    45m
  5. Lab: a neural network vs your best model
    Practical90m
  6. Lab: similarity with sparse and dense vectors
    Practical75m
  7. Machine learning project: build, evaluate and hand off
    Project300m

08Week 8 — Language Models, Prompts & Output Contracts

  1. How language models work, practically
    50m
  2. Hallucination, verification, cost and local vs hosted
    50m
  3. Calling a model from Python, provider-neutrally
    50m
  4. Prompts and output contracts
    55m
  5. Lab: the same questions with and without context
    Practical90m
  6. Lab: a narrow extraction task with schema validation
    Practical90m
  7. Weekly project: context report and versioned prompt
    Project180m
  8. Using model APIs responsibly: data, cost and NDPA
    25m

09Week 9 — Evaluating AI Systems

  1. Evaluation starts with success criteria
    50m
  2. Automated checks and human review
    50m
  3. Failure categories, regressions and trade-offs
    50m
  4. What an evaluation does and does not prove
    45m
  5. Lab: build a 20-case evaluation harness
    Practical90m
  6. Lab: compare two prompt versions and review regressions
    Practical90m
  7. Weekly project: an evaluation report
    Project180m

10Week 10 — Retrieval & Grounded Answers

  1. Retrieval: answering from your documents
    50m
  2. Chunking and a TF-IDF index in code
    55m
  3. Grounded answers with citations
    50m
  4. Retrieval failure vs generation failure
    50m
  5. Lab: index a small authorised collection
    Practical90m
  6. Lab: evaluate retrieval and answers separately
    Practical90m
  7. Weekly project: a grounded Q&A over your documents
    Project180m

11Week 11 — Tools, Agents & the AI Project

  1. Tools and agents: letting a model ask for data
    55m
  2. The agent loop, validation and limits
    55m
  3. Application reliability: configuration, diagnostics, safe failure
    50m
  4. Scoping a bounded assistant
    45m
  5. Lab: a read-only tool, tested five ways
    Practical90m
  6. Lab: evaluate the agent end to end
    Practical90m
  7. AI engineering project: a bounded, evaluated assistant
    Project300m

12Week 12 — Integrated Capstone

  1. From question to recommendation: scoping the capstone
    50m
  2. Lab: build sprint with traceable metrics
    Practical120m
  3. Lab: 10-minute presentation and defence rehearsal
    Practical90m
  4. Career pathways: analyst, ML engineer, AI engineer
    40m
  5. Integrated capstone: decision problem to defended recommendation
    Project480m
  6. Defending AI-assisted work in your capstone
    20m

Already enrolled? Sign in with your enrollment email to unlock all lessons.