South 고등학교 2학년 Foundations of Artificial Intelligence
Chapters: 4
1. AI and intelligent reasoning
How AI makes judgements · Search in AI · Blind vs informed search · Problems needing intelligent search · Knowledge representation
- Artificial Intelligence: How Machines Learn to Think – Artificial intelligence (AI) is the skill of a computer system to do tasks that normally need human thinking: seeing, understanding speech, deciding and learning. An AI system is an agent that senses, thinks and acts. Old AI followed rules written by people. Modern AI mostly uses machine learning: it finds its own rule from many labelled examples (data). Neural networks are layers of simple units whose link strengths (weights) change during training. AI is used in maps, translation, health, farming and games. It can be wrong or unfair when its data is one-sided (bias), so people must check it, protect privacy and stay responsible.
- Search Algorithms in AI: Finding the Way Step by Step – Many AI problems are about finding a way from a start to a goal: a route on a map, a puzzle solution or a good move in a game. We describe the problem as a state space: states (situations) joined by actions (moves). A search algorithm explores this space in a fixed order. Blind (uninformed) search like breadth-first and depth-first search knows nothing about where the goal is. Informed search uses a heuristic, a smart guess of how far the goal is, so it opens far fewer states. A* adds the cost so far (g) to the guess (h) and, with a heuristic that never over-guesses, still finds the shortest path.
2. AI and learning
Defining ML problems · Preparing data and features · Choosing ML models · Training and evaluation · Neural networks and deep learning · Deep learning applications
- Machine Learning – Machine learning (ML) is a way for computers to learn a rule from examples instead of being told the rule. We give data made of features (inputs) and, in supervised learning, labels (answers). The machine fits a model: regression predicts a number, a decision tree asks yes/no questions to pick a class, and k-means clustering groups unlabelled data. We train on most of the data, test on data it never saw, and measure accuracy or error. A model that only memorises (overfits) fails on new data.
- Neural Networks and Deep Learning – A neural network is a computer model made of many small units called artificial neurons. Each neuron multiplies its inputs by weights, adds them with a bias and passes the sum through an activation function. Neurons are arranged in layers: input, hidden and output. The network learns by training: it makes a guess, measures the error with a loss function, and changes the weights backwards (backpropagation with gradient descent) so the error becomes smaller. Networks with many hidden layers are called deep learning. CNNs are good at images, RNNs at sequences like text and speech.
3. Social impact of AI
AI and social change · Humans alongside AI · Critical view of AI · Ethics in AI use
- AI Ethics: Using Artificial Intelligence Fairly and Safely – AI systems learn patterns from data and then make decisions. AI ethics asks whether those decisions are fair, safe and respectful. The main issues are bias (unfair data gives unfair results), privacy (personal data needs consent and protection), transparency (people should know why an AI decided something), accountability (a human stays responsible), safety and misuse (deepfakes, false information), and social impact (jobs, the digital divide, the environment). Responsible AI means checking all of these before and after an AI is used.
4. AI project
AI for the SDGs · Designing an AI solution · Building the project · Assessing social impact
- The Sustainable Development Goals (SDGs): 17 Goals for 2030 – In 2015 all United Nations member countries adopted the 2030 Agenda with 17 Sustainable Development Goals (SDGs) and 169 targets. They cover people (poverty, hunger, health, education, equality), planet (water, climate, oceans, land), prosperity (energy, jobs, industry, cities), peace and partnership. All 17 are linked: progress in one helps the others.
- Machine Learning – Machine learning (ML) is a way for computers to learn a rule from examples instead of being told the rule. We give data made of features (inputs) and, in supervised learning, labels (answers). The machine fits a model: regression predicts a number, a decision tree asks yes/no questions to pick a class, and k-means clustering groups unlabelled data. We train on most of the data, test on data it never saw, and measure accuracy or error. A model that only memorises (overfits) fails on new data.
- AI Ethics: Using Artificial Intelligence Fairly and Safely – AI systems learn patterns from data and then make decisions. AI ethics asks whether those decisions are fair, safe and respectful. The main issues are bias (unfair data gives unfair results), privacy (personal data needs consent and protection), transparency (people should know why an AI decided something), accountability (a human stays responsible), safety and misuse (deepfakes, false information), and social impact (jobs, the digital divide, the environment). Responsible AI means checking all of these before and after an AI is used.