Definition and core purpose of deep learning

Summary

Deep learning is a branch of machine learning that uses multilayered neural networks to transform input data into progressively more abstract and useful representations. It is commonly used for tasks such as classification, regression, and representation learning. The “deep” in deep learning refers to the use of many layers (from a few to hundreds or thousands), which creates a long chain of transformations from input to output. A core purpose of deep learning is to learn features automatically from data rather than relying on hand-crafted feature engineering. For example, in image recognition, early layers can learn simple patterns (like edges), while later layers learn more complex structures (like parts of faces) and ultimately the target concept (like a face). Deep learning models can be trained in supervised, semi-supervised, or unsupervised ways, enabling them to learn from both labeled and abundant unlabeled data.

Key Points

• 1

  Deep learning uses multilayer neural networks to convert inputs into increasingly abstract representations.

• 2

  Its core purpose is automatic feature learning from data, reducing the need for hand-crafted feature engineering.

• 3

  The term “deep” refers to many layers, creating a deep chain of transformations (credit assignment path) from input to output.

• 4

  Deep learning models can be trained using supervised, semi-supervised, or unsupervised learning, including learning from unlabeled data.