Welcome to AI Concepts
Artificial Intelligence (AI) might seem like magic, but it's actually just math and logic working together to solve problems. This guide breaks down complex topics into simple explanations with interactive demos.
How to use this guide:
- Read the simple explanation.
- Relate it to the real-world example.
- Play with the interactive demo to see it in action.
Select a topic from the menu to get started! Each topic also has a full simulation in the AI Simulations tab.
K-Means Clustering
What is it? K-Means is an algorithm used to group data into categories (clusters) without being told what the categories are. It finds the "center" of groups automatically.
Demo: Grouping the Dots
Click "Step" to see how the computer finds the centers (X) of the colored groups.
Status: Random Centers
K-Nearest Neighbors (k-NN)
What is it? k-NN is a classification method. It assumes that similar things exist in close proximity. To predict what a new data point is, it looks at its closest neighbors.
Demo: Classify the Mystery Point
The is a mystery. Look at its 5 nearest neighbors.
Neural Networks
What is it? Inspired by the human brain, a neural network consists of layers of "neurons". Each connection has a "weight" (importance). Inputs are multiplied by weights to produce an output.
Input 1: Is it cold? (High weight/importance)
Input 2: Is it sunny? (Low weight/importance)
If the total score crosses a threshold, the "Wear Jacket" neuron fires.
Demo: Single Neuron
Adjust the Weight. If Input × Weight > 50, the neuron fires!
⚡
(Threshold: 50)
A* Pathfinding
What is it? A* (A-Star) is an algorithm used to find the shortest path between two points. It is smart because it uses a "heuristic" (an educated guess) to prioritize paths that move towards the goal, rather than exploring blindly.
Demo: Blind Search vs. A* Search
Imagine searching for a key in a dark room.
Blind Search (Dijkstra)
Checks every direction equally.
A* Search
Guesses direction towards Goal.
Logic Gates
What is it? Logic gates are the building blocks of computers. They take binary inputs (On/Off, 1/0) and produce a single output based on a rule.
AND Gate: Security Box. You need Key A AND Key B to open it.
OR Gate: Doorbell. Front button OR Back button rings the bell.
Demo: Circuit Builder
Reinforcement Learning (Q-Learning)
What is it? A type of Machine Learning where an agent learns by trial and error. It gets a "Reward" for doing something good and a "Penalty" for doing something bad.
Dog sits → Give Treat (+ Reward). Dog learns to sit.
Dog chews shoe → Say "No!" (- Penalty). Dog learns not to chew shoes.
Demo: Train the Robot
The robot wants to reach the Battery 🔋.
Robot Action: Waiting
Robot Confidence: 0%