Artificial intelligence, machine learning, and deep learning are related ideas, but they are not the same thing. AI is the broad goal of making computers perform tasks that seem intelligent, such as planning, recognizing speech, or playing games. Machine learning is one way to build AI by training systems with data instead of writing every rule by hand.
Deep learning is a special kind of machine learning that uses layered neural networks to find complex patterns.
Key Facts
- AI is the largest category: it includes rule based systems, search algorithms, planning, robotics, machine learning, and deep learning.
- Machine learning is a subset of AI: a model learns a pattern from data so it can make predictions or decisions on new examples.
- Deep learning is a subset of machine learning: it uses neural networks with many layers to learn features from data.
- A simple supervised learning idea is prediction = model(input), where the model is adjusted using examples with known answers.
- A common error measure is mean squared error: MSE = (1/n) Σ(y - ŷ)^2.
- Deep learning often needs large datasets and strong computing power, while simpler machine learning can work well with smaller, structured datasets.
Vocabulary
- Artificial Intelligence
- Artificial intelligence is the broad field of making computers perform tasks that normally require human-like reasoning, perception, or decision making.
- Machine Learning
- Machine learning is a method of building AI systems that improve their performance by finding patterns in data.
- Deep Learning
- Deep learning is a type of machine learning that uses neural networks with multiple layers to learn complex patterns.
- Model
- A model is a mathematical or computational system trained to turn inputs into useful outputs such as labels, numbers, or decisions.
- Training Data
- Training data is the collection of examples used to teach a machine learning model how inputs relate to correct outputs.
Common Mistakes to Avoid
- Calling every computer program AI is wrong because many programs simply follow fixed instructions without learning, reasoning, or adapting.
- Using AI, machine learning, and deep learning as exact synonyms is wrong because machine learning is inside AI, and deep learning is inside machine learning.
- Assuming deep learning is always the best choice is wrong because simpler methods can be faster, easier to explain, and more accurate when the dataset is small or structured.
- Thinking a trained model understands like a human is wrong because it usually detects statistical patterns and can fail on examples that differ from its training data.
Practice Questions
- 1 A school has 1,000 labeled emails, and 120 are spam. What percentage of the dataset is spam?
- 2 A machine learning model makes 200 image predictions and gets 170 correct. What is its accuracy as a percent?
- 3 A city wants software that follows a fixed list of traffic light rules, and another system that learns to predict traffic jams from years of sensor data. Which one is better described as traditional programming, and which one is machine learning? Explain your reasoning.