Science focus: concepts are explained with scientific and research-oriented examples. Questions are optional and never block the next unit.
TOPIC 1

Machine Learning and Data-Driven Intelligence

Machine learning (ML) is an approach in which algorithms learn useful patterns from data rather than relying only on explicitly written rules.

Science example
A model can learn a relationship between environmental measurements and an observed outcome.
Quick revision: Machine Learning and Data-Driven Intelligence is a key concept to be able to define, explain and apply in a simple scientific context.
TOPIC 2

Supervised Learning

Supervised learning uses examples with known target labels or values. It commonly supports classification and regression.

Science example
Classifying plant leaves as healthy/diseased from labelled images.
Quick revision: Supervised Learning is a key concept to be able to define, explain and apply in a simple scientific context.
TOPIC 3

Unsupervised Learning

Unsupervised learning works with data without target labels and seeks structure such as groups or lower-dimensional representations.

Science example
Grouping water samples according to similar chemical measurements.
Quick revision: Unsupervised Learning is a key concept to be able to define, explain and apply in a simple scientific context.
TOPIC 4

Reinforcement Learning

An agent learns actions through interaction with an environment using rewards or penalties.

Science example
A simulated robot can learn a movement strategy by receiving rewards for reaching a target.
Quick revision: Reinforcement Learning is a key concept to be able to define, explain and apply in a simple scientific context.
TOPIC 5

Training Data and Labelled Data

Training data are examples used to fit a model. Labelled data include the correct class or target value associated with each example. Data quality strongly affects results.

Science example
Microscope images labelled by experts can be used to train a classifier.
Quick revision: Training Data and Labelled Data is a key concept to be able to define, explain and apply in a simple scientific context.
TOPIC 6

Regression and Classification

Regression predicts continuous numerical values; classification predicts categories.

Science example
Regression: predict temperature. Classification: identify a species category.
Quick revision: Regression and Classification is a key concept to be able to define, explain and apply in a simple scientific context.
TOPIC 7

Clustering and Dimensionality Reduction

Clustering groups similar observations. Dimensionality reduction represents high-dimensional data using fewer informative dimensions for exploration or modelling.

Science example
Cluster gene-expression profiles; reduce many measured variables for visual exploration.
Quick revision: Clustering and Dimensionality Reduction is a key concept to be able to define, explain and apply in a simple scientific context.
TOPIC 8

Neural Networks

A neural network consists of connected computational units arranged in layers. During training, connection parameters are adjusted to reduce prediction error.

Science example
Neural networks are widely used for scientific image and signal analysis.
Quick revision: Neural Networks is a key concept to be able to define, explain and apply in a simple scientific context.
TOPIC 9

Decision Trees

A decision tree makes predictions through a sequence of feature-based decision rules arranged like branches.

Science example
Classify a sample by testing measured properties at successive decision nodes.
Quick revision: Decision Trees is a key concept to be able to define, explain and apply in a simple scientific context.
TOPIC 10

k-Nearest Neighbours (k-NN)

k-NN predicts using the labels or values of the most similar nearby training examples according to a distance measure.

Science example
A new specimen may be classified from similar measured specimens.
Quick revision: k-Nearest Neighbours (k-NN) is a key concept to be able to define, explain and apply in a simple scientific context.
TOPIC 11

Deep Learning

Deep learning uses neural networks with multiple processing layers and is especially useful for complex data such as images, audio and text.

Science example
Deep models can analyse medical or satellite images, with appropriate validation and oversight.
Quick revision: Deep Learning is a key concept to be able to define, explain and apply in a simple scientific context.
TOPIC 12

Model Training and Evaluation

A model is trained on data and evaluated on data not used for fitting. Accuracy is the fraction of predictions that are correct; precision measures how many predicted positives are actually positive; recall measures how many actual positives are found.

Science example
In disease screening, recall can be important when missing a true positive has serious consequences.
Quick revision: Model Training and Evaluation is a key concept to be able to define, explain and apply in a simple scientific context.
TOPIC 13

Real-Life ML Applications

ML is used for prediction, classification, anomaly detection, recommendation and pattern discovery across science and society.

Science example
Examples include weather-related prediction, image analysis, sensor monitoring and biological data analysis.
Quick revision: Real-Life ML Applications is a key concept to be able to define, explain and apply in a simple scientific context.

End of Unit 2

You may practice now or skip directly to the next unit.

Practice Questions & AnswersSelf-check QuizSkip & Study Unit 3 →