AI & Machine Learning
Learn how AI systems learn from data: core algorithms, training models and evaluating results.
- Fees
- Contact us for fees
About this course
This course teaches how machines learn from data, and how to turn a trained model into something people can use. It goes further into modelling than the Data Science course: classical machine learning, neural networks, images and text, and then serving a model behind an API.
You work in Python, from a short refresher through scikit-learn, PyTorch or TensorFlow, and Hugging Face. Every model is judged on data it has not seen, and fairness and responsible use are discussed alongside accuracy.
What you will be able to do
- Frame a problem as classification, regression, clustering or ranking, and choose a sensible baseline.
- Prepare data without leaking test information into training.
- Train and compare linear models, tree ensembles and gradient boosting, and tune them with cross-validation.
- Build, train and evaluate a neural network for images or text.
- Use pretrained models from Hugging Face for text and image tasks.
- Track experiments, serve a model through an API and describe how to monitor it after launch.
- Spot bias, data drift and other risks, and explain them to a non-specialist.
Syllabus
10 modules. Select a module to see its topics.
Module 1: AI and machine learning foundations
- AI, machine learning and deep learning compared
- Types of learning
- The machine learning workflow
- Framing a problem
- Data and bias
- Setting up Python and notebooks
Module 2: Python and maths for machine learning
- NumPy and pandas
- Plotting
- Vectors and matrices
- Gradients and optimisation in plain language
- A probability refresher
Module 3: Data preparation and feature engineering
- Cleaning
- Encoding categories
- Scaling
- Train, validation and test splits
- Avoiding data leakage
- scikit-learn pipelines
Module 4: Regression
- Linear and polynomial regression
- Regularisation (ridge and lasso)
- Error metrics
- Residual analysis
Module 5: Classification
- Logistic regression
- k-nearest neighbours
- Decision trees and random forests
- Gradient boosting
- Support vector machines
- Confusion matrix, precision, recall and ROC-AUC
- Imbalanced classes
Module 6: Unsupervised learning
- k-means and hierarchical clustering
- PCA
- Anomaly detection
- Recommender-system basics
Module 7: Model evaluation and tuning
- Cross-validation
- Hyperparameter search
- The bias-variance trade-off
- Overfitting
- Ensembles
- Explaining model decisions
Module 8: Deep learning fundamentals
- Neurons and layers
- Activation functions
- Backpropagation
- Training loops in PyTorch or TensorFlow with Keras
- Regularisation
- Convolutional networks for images
Module 9: Language and sequence models
- Text preprocessing
- Embeddings
- Recurrent networks in outline
- The transformer idea
- Hugging Face pipelines
- Text classification and sentiment analysis
Module 10: From model to product
- Saving and serving a model with FastAPI
- Docker basics
- Experiment tracking with MLflow
- Monitoring and drift
- Fairness and responsible AI
- Capstone project
Tools and technologies
- Python
- NumPy
- pandas
- scikit-learn
- XGBoost or LightGBM
- PyTorch or TensorFlow with Keras
- Hugging Face Transformers
- OpenCV (introduction)
- MLflow
- FastAPI
- Docker
- Jupyter
- Git and GitHub
Projects you will build
House price prediction with regression and error analysis.
Customer churn classifier that handles imbalanced classes and explains its decisions.
Customer segmentation with clustering.
Image classifier built with a convolutional neural network.
Sentiment analysis of product reviews with a pretrained transformer.
Capstone: a trained model served through an API, with a short monitoring plan and a model card describing its limits.
Before you start
- This outline starts from the basics: Python and the maths used in machine learning are covered in the first modules.
- Helpful but not required: school-level maths and some programming exposure.
- Entry requirements, the lab set-up and any cloud costs are confirmed by Pathway Career when you enquire.
Career paths
Examples of roles that use these skills. They are not a promise of a job.
- Junior machine learning engineer
- AI engineer
- Data scientist
- NLP engineer
- Computer-vision engineer
- MLOps associate
- Research assistant
Questions about this course
What maths do I need?
School-level maths is enough to begin. Linear algebra, calculus and probability ideas are introduced as they are needed. Entry requirements are confirmed by Pathway Career when you enquire.
How is this different from Data Science?
Data Science stresses analysis, statistics and communication, with machine learning as one part. This course goes further into modelling, neural networks, language and images, and deployment.
How is this different from Generative AI?
Here you train and evaluate models yourself. The Generative AI course builds applications on ready-made language models.
PyTorch or TensorFlow?
One framework is used for the main exercises, and the ideas carry over to the other. Pathway Career confirms which one the course uses when you enquire.
Do I need a powerful computer?
Not necessarily. Heavy training can use free or low-cost cloud notebooks. Pathway Career confirms the arrangement when you enquire.
Is this a good next step after the Python course?
Yes. The Python course builds the programming base, and this course adds modelling on top of it.
Next step
Ready to learn how models are built and put to use? Send an enquiry and ask for the full AI & Machine Learning syllabus, schedule and fees.
Send an enquiry