Skip to main content
Practice

What is Deep Learning?

In simple terms, Deep Learning is a technology that enables computers to learn like humans, and it's a subfield of Machine Learning.

In deep learning, computers learn from massive amounts of data and use the learned results to predict or classify new data.

The core of deep learning is the structure known as an Artificial Neural Network. An artificial neural network mimics how the human brain processes information through a computer program.


Key Concepts of Deep Learning​

  1. Neuron: In deep learning, small processing units work together, called 'neurons'. These are modeled after the nerve cells in the human brain.

  2. Layer: Neurons are organized into layers, typically divided into Input Layer, Hidden Layer, and Output Layer. The input layer receives the data, the hidden layer processes the data, and the output layer produces the final result.

  3. Learning: Deep learning models learn through a vast amount of data. For example, if you show numerous cat photos, the model learns the characteristics of a cat. Later, when new photos are shown, it can predict whether it is a cat or not.


Deep learning is noted for its ability to automatically extract features from data, so there is no need for humans to explicitly define the features of the data.

Additionally, a large amount of data is required for AI model training, and many computations are performed during the learning process.


Example of Deep Learning​

If you create a program to recognize handwritten digits using deep learning, it would follow a process like this:

  1. Data Preparation: Prepare thousands of handwritten digit images. Each image includes a digit from 0 to 9 and comes with a label (answer) indicating what the digit is.

  2. Model Training: Input these images into the deep learning model and tell the model what digit each image represents. The deep learning model analyzes the pixel patterns of the images to automatically extract and learn the features of the digits. The model autonomously discovers important features without human intervention.

  3. Prediction: When a new handwritten digit image is input into the trained AI model, the AI model predicts which digit the image represents. The AI model uses the learned features from the training data to make predictions on new data.

Want to learn more?

Join CodeFriends Plus membership or enroll in a course to start your journey.