NCA-AIIO - Essential AI Knowledge (38% of the exam) - Section 1.3
Differentiate the concepts of AI, machine learning, and deep learning.
Define artificial intelligence as the broad field of enabling machines to perform intelligent tasks, machine learning as the subset that learns patterns from data, and deep learning as the neural-network-driven subset of ML. Recognise that deep learning's reliance on large models and datasets is the primary reason GPU acceleration is essential.
Practice question for this objective
lock_openFree sampleEssential AI Knowledgeeasy
Which statement most accurately describes the relationship between artificial intelligence, machine learning, and deep learning?
AThey are three independent and equally broad disciplines that overlap only in their use of statistical methods
BArtificial intelligence is the broadest field; machine learning is a subset of AI that learns from data; deep learning is a subset of machine learning that uses multi-layer neural networkscheck_circle Correct
CDeep learning is the broadest category, with machine learning and artificial intelligence as progressively narrower specialisations within it
DMachine learning and deep learning are parallel branches of artificial intelligence that share no techniques or tools
Identify the correct nesting relationship in which machine learning is a subset of AI and deep learning is a subset of machine learning. The three disciplines sit in a strict nesting arrangement often illustrated as concentric circles. Artificial intelligence is the outermost circle, covering any computational approach that mimics human intelligence, including rule-based expert systems, search algorithms, and learning methods. Machine learning is a subset that focuses specifically on systems that improve their performance on a task through exposure to data, without being explicitly programmed for every case. Deep learning sits innermost, restricting ML further to models built from neural networks that contain many hidden layers, enabling automatic extraction of hierarchical features from raw data.
Why A is wrong: This contradicts the established nesting relationship. ML is a proper subset of AI, and DL is a proper subset of ML; they are not three co-equal, independent disciplines.
Why B is correct: This describes the canonical concentric-circles model: AI encompasses all techniques that exhibit human-like reasoning, ML refines that to data-driven learning, and DL further specialises to neural networks with multiple hidden layers.
Why C is wrong: This inverts the correct hierarchy. Deep learning is the narrowest and most specialised of the three; it sits inside machine learning, which in turn sits inside artificial intelligence.
Why D is wrong: Deep learning is a subset of machine learning, not a parallel branch. They share training data, gradient-based optimisation, and evaluation practices, so the claim that they share no techniques is incorrect.
Answers that look right on this material and are not. Each one is a distractor from a different question in the NCA-AIIO bank for this domain.
Deep learning is the broadest field, machine learning sits inside it, and artificial intelligence is the narrowest specialisation.
errorWhy it is wrong: Tempting because deep learning is the most advanced of the three and advanced can feel like broadest, but the nesting is the reverse: deep learning is the narrowest subset, not the containing field.
Deep neural networks contain fewer parameters than classical models, leaving no room to store hand-crafted features.
errorWhy it is wrong: Tempting because parameter count sounds relevant, but deep networks typically have far more parameters than classical models, not fewer, and parameter count is not why feature engineering is skipped.
It is an example of machine learning but not of artificial intelligence, because it competes at a board game.
errorWhy it is wrong: Tempting because game-playing feels like learning, but competing at a game does not make a system machine learning when it never learns from data, and machine learning is itself a part of AI rather than separate from it.