Applied Machine Learning
Go from classic ML to the models used at work: neural networks, embeddings, CNNs, gradient boosting and LLM apps, plus evaluation, experiment tracking and a capstone.
Leads to: Data Scientist, Machine Learning Engineer
Before you start
This course builds on Machine Learning Foundations. Take it first if you are new to the topic.
About this course
A mid-level course for data scientists and machine learning engineers who have finished Machine Learning Foundations. You will build and train a small neural network in NumPy, diagnose over- and underfitting, and learn the PyTorch and Keras vocabulary of tensors, models, losses and optimisers. You will classify text with embeddings, understand how CNNs and transfer learning handle images, and implement gradient boosting, the default strong model for tabular data. A module on large language models covers tokens, context windows and sampling, retrieval-augmented generation, and how to build an evaluation set for models and LLM apps. The course ends with the habits of professional practice: experiment tracking with MLflow, model cards that explain results to non-specialists, and an end-to-end capstone project for your portfolio. Code activities run in the browser with NumPy and plain Python; framework work (scikit-learn, PyTorch, Keras, sentence-transformers, MLflow) is taught in the reading and practised in a free local Jupyter, Colab or Kaggle notebook. Prerequisite: Machine Learning Foundations.
Course content
Module 1
Neural Networks
3 lessons · 2h 4mBuild and train a small neural network by hand, diagnose over- and underfitting, and learn how PyTorch and Keras express the same ideas.
- 51mNeural network basicsReading
Build a two-layer neural network in NumPy, write its backward pass and train it to solve XOR.
- 47mTraining deep modelsReading
Read training and validation curves to diagnose over- and underfitting, and fix them with regularisation and early stopping.
- 26mPyTorch or Keras tourReading
Learn how PyTorch and Keras express tensors, models, losses, optimisers and the training loop.
Module 2
Common Domains
3 lessons · 2h 7mApply ML to text with embeddings, to images with CNNs and transfer learning, and to tabular data with gradient boosting.
- 44mText and embeddingsReading
Turn text into embedding vectors, compare them with cosine similarity and classify support messages.
- 22mImagesReading
Understand how convolutional networks find patterns in images and how transfer learning reuses pretrained models.
- 1h 1mGradient boostingReading
Implement gradient boosting with decision stumps and learn how XGBoost, LightGBM and scikit-learn apply it to tabular data.
Module 3
Working With LLMs
3 lessons · 2h 28mUnderstand how large language models work, ground them in your documents with retrieval-augmented generation, and evaluate models and LLM apps with a fixed test set.
- 19mHow LLMs workReading
Explain how large language models predict tokens, what the context window limits, and how sampling settings change outputs.
- 1h 19mRetrieval-augmented generationReading
Design a RAG pipeline that retrieves relevant passages from your documents and grounds an LLM's answers in them.
- 50mEvaluating models and LLM appsReading
Build a fixed evaluation set with references and tags, and score a model or LLM app with exact match and token F1.
Module 4
Projects
3 lessons · 3h 25mTrack experiments, communicate model results with a model card, and deliver an end-to-end capstone project.
- 1h 42mExperiment trackingReading
Log parameters, metrics, artifacts and context for every training run with MLflow, and compare runs to pick a model.
- 59mCommunicating model resultsReading
Write a model card and explain a model's results, limits and risks in terms a decision-maker can act on.
- 44mCapstone projectReading
Plan, build, evaluate and document an end-to-end machine learning project for your portfolio.