Click image to open full size
What Is AI?
AI vs Machine Learning vs Deep Learning
Related Tools
Related Labs
Artificial intelligence, machine learning, and deep learning are closely related ideas in computer science, but they are not the same thing. Artificial intelligence, or AI, is the broad goal of making computers perform tasks that seem to require human intelligence. Machine learning, or ML, is one way to achieve AI by letting systems learn patterns from data instead of following only fixed rules. Deep learning is a more specialized form of machine learning that uses layered neural networks to model complex patterns.
The relationship is often shown as nested sets: deep learning inside machine learning, and machine learning inside artificial intelligence. Traditional AI can include rule based systems, search algorithms, and logic, even when no learning happens. Machine learning usually depends on training data and improves performance by adjusting model parameters. Deep learning is especially useful for images, speech, and language because many network layers can automatically learn useful features from raw input.
Key Facts
- AI is the broad field of building systems that perform intelligent tasks such as reasoning, planning, perception, and decision making.
- ML is a subset of AI where a model learns from data instead of being programmed with every rule explicitly.
- Deep learning is a subset of ML that uses multilayer neural networks, often written as y = f(x; theta).
- A common ML goal is to minimize prediction error, for example Loss = (1/n) * sum((y - y_hat)^2).
- In supervised learning, the model maps inputs to labeled outputs, written as y_hat = f(x).
- More data and larger models often help deep learning, but they also increase computation time and energy use.
Vocabulary
- Artificial Intelligence
- The broad area of computer science focused on creating systems that can perform tasks associated with human intelligence.
- Machine Learning
- A branch of AI in which computers learn patterns from data to make predictions or decisions.
- Deep Learning
- A type of machine learning that uses neural networks with many layers to learn complex patterns.
- Training Data
- The examples a model uses to learn relationships between inputs and outputs.
- Neural Network
- A computational model made of connected units that transform input data through layers to produce an output.
Common Mistakes to Avoid
- Saying AI, ML, and deep learning all mean the same thing, which is wrong because they describe different levels of a hierarchy with different methods and scope.
- Assuming every AI system learns from data, which is wrong because some AI systems use fixed rules, logic, or search without machine learning.
- Thinking deep learning always performs best, which is wrong because it often needs large datasets, high computing power, and may be unnecessary for simple problems.
- Believing more training data automatically guarantees a fair and accurate model, which is wrong because biased or low quality data can produce biased or unreliable results.
Practice Questions
- 1 A school uses a rule based program that follows fixed if then statements to schedule classrooms. Is this AI, ML, deep learning, or more than one of these? Explain your choice.
- 2 A model is tested on 4 examples with true values y = [3, 5, 2, 6] and predictions y_hat = [2, 5, 4, 7]. Calculate the mean squared error using Loss = (1/n) * sum((y - y_hat)^2).
- 3 A company wants to identify objects in millions of photos. Explain why deep learning might be preferred over a simple rule based system, and give one tradeoff.