Teacher Resources (new)
A 4.4.1 / A 4.4.2 Ethics of Emerging Tech
A4.4.1 discusses ethical implications of ML in real-world scenarios: accountability, fairness, bias (especially from training data), consent, environmental/societal impact, privacy, security, and transparency,...
Teacher Resources (new)
A 4.3.10 / A 4.4.1 Model Selection & Ethical Implications.
A4.3.10 emphasises the importance of model selection and comparison in machine learning. Different algorithms yield varied results based on data characteristics and problem types. Selecting specific models...
Teacher Resources (new)
A 4.3.9 Convolutional Neural Networks (CNNs)
A4.3.9 describes how Convolutional Neural Networks (CNNs) adaptively learn spatial hierarchies of features in images. Their basic architecture includes input, convolutional, activation, pooling, fully...
Teacher Resources (new)
A 4.3.8 Artificial Neural Networks (ANNs)
A4.3.8 outlines Artificial Neural Networks (ANNs), which simulate interconnected "neurons" to process and learn from data. These networks enable tasks like classification, regression, and pattern recognition....
Teacher Resources (new)
A 4.3.6 / A 4.3.7 Reinforcement Learning & Genetic Algorithms
A4.3.6 covers reinforcement learning, where agents learn decisions by interacting with their environment, based on cumulative reward, actions, states, rewards, policies, and the exploration vs. exploitation...
Teacher Resources (new)
A 4.3.4 / 4.3.5 Unsupervised Learning: Clustering & Association
A4.3.4 details unsupervised learning's clustering techniques, which group data points based on inherent feature similarities. A practical application is segmenting a customer base using purchasing data....
Teacher Resources (new)
A 4.3.3 Model Evaluation & Hyperparameter Tuning
A4.3.3 explains the role of hyperparameter tuning in optimising supervised learning algorithms. It details crucial evaluation metrics like accuracy, precision, recall, and the F1 score (calculated as...
Teacher Resources (new)
A 4.3.2 Supervised Learning - Classification
A4.3.2 focuses on understanding how machine learning classification techniques are utilised to predict "discrete categorical outcomes". Unlike linear regression, which predicts continuous numerical values...
Teacher Resources (new)
A 4.3.1 Linear regression
4.3.1 explains linear regression, a Higher Level (HL) machine learning approach, used to predict continuous outcomes. This technique focuses on the relationship between independent (predictor) and dependent...
Teacher Resources (new)
A 4.2.3 Dimensionality Reduction
A4.2.3 describes the importance of dimensionality reduction to mitigate the "curse of dimensionality". This "curse" involves issues like overfitting, high computational complexity, data sparsity, impact...