Think Together · Start with a Short Story
The girl hears “deep learning” and thinks the robot is especially serious. Here, “deep” mainly refers to the model's layers.
Lay Out the Process
- Machine learningImprove from data
- Neural networkA computing structure with adjustable parts
- Many layersLearn and transform step by step
Deep learning usually means using a neural network with many layers to learn representations and perform tasks.
Layered structures can handle complex data such as images, sounds, and language, but results depend on the design, data, training, and computing resources.
“Deep” does not mean profound thinking, and it does not automatically make the method better for every task.
Now You Try
To check whether one integer is greater than 10, do you need to train a deep model?
Ready? Open this to see the explanation
No. One clear comparison rule can do it accurately. Training would only add complexity.
One Idea to Take Away
Deep learning usually means using a neural network with many layers to learn representations and perform tasks.
Try covering the diagram above and explaining each step to someone in your family. If one step is hard to explain, return to that card and look again.