Video transcript
Supervised learning
How does an AI system learn through supervised learning? It all starts with data, such as these images.
At first, the images are unlabeled. If a developer provided the AI system with these images in this state,
the AI wouldn’t be able to learn from them. So, they must be labeled.
The images are labeled. This one is a car. This one is a bus. This one is a bicycle. Once labeled, the
examples are organized so the AI can recognize patterns. After seeing many labeled examples, the AI
starts to learn. It begins to identify features that make something a car, bus, or bicycle, such as shape,
wheels, and windows.
Eventually, the AI builds a model based on what it has observed and learned. Now, when a new image,
like this one, is presented, it can offer insights. In this case, it confidently says, “I’ve identified this as a
car.”
Page 1 of 1