AI, graphs and neural networks
| English | Chinese | Pinyin |
|---|---|---|
| machine learning | 机器学习 | jī qì xué xí |
| deep learning | 深度学习 | shēn dù xué xí |
| neural networks | 神经网络 | shén jīng wǎng luò |
| graph | 图 | tú |
| hidden layers | 隐藏层 | yǐn cáng céng |
| nodes | 节点 | jié diǎn |
| edges | 边 | biān |
| minimax | 极小化极大 | jí xiǎo huà jí dà |
| weight | 权重 | quán zhòng |
| activation function | 激活函数 | jī huó hán shù |
Artificial intelligence
- AI builds systems that do tasks normally needing human intelligence — vision, speech, translation, driving.
- Most modern AI uses machine learning 机器学习, and within it deep learning 深度学习 with neural networks 神经网络.
- Many AI problems are searched on a graph 图.
What AI is
- Machine learning = algorithms that learn patterns from data instead of being programmed step by step.
- Deep learning uses neural networks with many layers — dominant since the 2010s.
- These let models learn straight from raw data (pixels, audio, text).
A neural network with an input layer, two hidden layers 隐藏层 and an output layer
Machine learning means an algorithm that:
ML learns from data; deep learning is ML using multi-layer neural networks.
Graphs in AI
- Many AI problems sit on a graph — nodes 节点 (states/places) joined by edges 边 (moves/relationships).
- Pathfinding: roads form a graph; shortest route via Dijkstra or A*.
- Game playing: each position is a node, each move an edge; minimax 极小化极大 searches the game tree.
- Knowledge representation: a semantic network has concepts as nodes and relationships as edges. BFS and DFS are standard graph-search tools.
A self-driving car uses cameras and lidar sensors to see the road around it
In AI, a graph consists of:
Graphs model states/places as nodes and moves/relationships as edges — used for pathfinding, game trees and knowledge.
Artificial neural networks
- An artificial neuron multiplies each input by a weight 权重, adds them (with a bias), applies a non-linear activation function 激活函数, and outputs the result.
- Neurons sit in an input layer, one or more hidden layers (where internal patterns are learned), and an output layer.
- Many hidden layers → a deep neural network; training it is deep learning.
An artificial neuron multiplies each input by a weight, sums them with a bias, then applies an activation function
- Pathfinding uses Dijkstra's algorithm and the A* algorithm; neural networks train by back propagation of errors; regression methods predict numbers.
Tap the parts of a neural network
Explore the layers. Data flows left to right: the input layer takes the features, the hidden layers learn patterns, and the output layer gives the answer — with every connection carrying a weight that training adjusts.
Match each part of a neural network to its role.
Features enter the input layer, hidden layers learn patterns, the output layer answers; learning = tuning the weights.
A neural network is called "deep" when it has many hidden layers, and each neuron takes a weighted sum of its inputs (plus a bias) before applying an activation function.
Stacking many hidden layers lets the network learn richer patterns — training such a network is deep learning.
A neural network is called "deep" when it has:
Many hidden layers make a deep neural network; training one is deep learning.
You've got it
- machine learning learns from data; deep learning uses multi-layer neural networks
- many AI problems are graph search (pathfinding, minimax game trees, semantic networks)
- a neuron = weighted sum + bias → activation function → output
- layers: input → hidden(s) → output; many hidden layers = deep