AI Dictionary › AI Fundamentals

Curriculum Learning

Curriculum learning is a training strategy in which data is presented to the model in a deliberate order, typically from simplest to most complex, rather than randomly. The idea is inspired by how humans learn, building skills progressively. The goal is to make training more efficient or to improve the final quality of the model.

Definition

How it works

Instead of mixing all training examples together from the start, curriculum learning defines a difficulty criterion and organizes the data into a progressive sequence. The model is first exposed to easier examples, to build solid basic representations, and only later to more difficult or ambiguous ones.

Applications

It has been applied in various neural network training contexts, including the pre-training of language models, where simple texts can precede more complex or specialized ones. It is also used in areas such as reinforcement learning, where tasks can be ordered by increasing difficulty.

History & etymology

It is an idea borrowed from pedagogical concepts applied to machine learning, proposed as a way to improve convergence speed and model generalization.

Related terms

• Pre-training
• Supervised Learning
• Reinforcement Learning
• Training Data
• Overfitting

More in AI Fundamentals

• LLM (Large Language Model)
• Token
• Context Window
• Temperature
• AI Hallucination
• Fine-tuning
• RAG (Retrieval Augmented Generation)
• Embedding

Put it into practice

• Real scenario: apply this concept with an instant AI verdict.
• Daily gym: short drills to train dictionary concepts.
• Playoff game: challenge other players, elimination-style.

📱 Download the Android app (beta) iOS coming soon

Say what you mean. Get what you need.

Grace Certified, the AI coach that trains and certifies your prompt engineering, by Agora Intelligence.