✦ScienexySCIENCE KNOWLEDGE LIBRARY
Science reference · SCI-CS-003-056

Supervised Learning

Supervised learning is a machine learning approach in which a model learns a relationship between inputs and known target outputs using labeled training examples.

fθ(x) ≈ y

Explanation

Supervised learning is one of the most widely used machine learning paradigms. It trains a model using a dataset in which each example contains input features and a corresponding target label or numerical value. The learning algorithm attempts to discover a mapping from inputs to outputs that performs well on previously unseen examples. Classification is a supervised learning task in which the target is a category, such as spam or not spam. Regression is a task in which the target is a numerical value, such as a house price or temperature. During training, a loss function measures the difference between the model's predictions and the provided targets. An optimization algorithm then adjusts model parameters to reduce this loss. Common supervised algorithms include linear regression, logistic regression, decision trees, random forests, support vector machines, and neural networks. The dataset is commonly divided into training, validation, and test subsets to support model fitting, model selection, and final evaluation. Supervised learning requires suitable target information, and its performance depends on label quality, data representativeness, and the model's ability to generalize.

Variables

SymbolMeaning

x

The input features supplied to the model.

y

The known target output or label.

fθ

The prediction function parameterized by θ.

θ

The model parameters learned during training.

Loss

A measure of the difference between predicted outputs and target values.

Example or application

A model is trained using house records containing floor area, location, and the actual selling price. It learns relationships between the input features and prices, then predicts the price of a house not included in the training dataset.

Text

In supervised learning, each training example includes an input and its corresponding target. The model learns from these examples so that it can predict outputs for new inputs.

Key Points

  • Supervised learning uses labeled training data.

  • Each training example contains an input and a corresponding target.

  • Classification predicts categories.

  • Regression predicts numerical values.

  • A loss function measures prediction errors.

  • Evaluation on unseen data helps measure generalization.

Applications

  • Email spam classification

  • House-price prediction

  • Disease classification from medical data

  • Credit-risk prediction

  • Image recognition

  • Demand forecasting

Learning Outcomes

  • Define supervised learning.

  • Explain the role of labeled datasets.

  • Distinguish classification from regression.

  • Identify common supervised learning algorithms.

  • Explain how a model learns from target outputs.

Practice Questions

  • What is supervised learning?

  • What is the difference between classification and regression?

  • Why are labels necessary in traditional supervised learning?

  • Name three supervised learning algorithms.

Related Concepts

  • Types of Machine Learning

  • Classification

  • Regression

  • Loss Functions

  • Model Evaluation

Common Mistakes

  • Confusing classification with regression.

  • Using incorrect or inconsistent labels during training.

  • Assuming high training accuracy guarantees strong test performance.

  • Evaluating a model only on examples it has already seen.

Sources and verification

Sources are shown as provided in the project data. Not every source has yet been independently verified. Please report errors or missing citations.

  • Bishop, C. M. Pattern Recognition and Machine Learning.
  • Hastie, T., Tibshirani, R., and Friedman, J. The Elements of Statistical Learning.
  • Murphy, K. P. Probabilistic Machine Learning: An Introduction.

Suggest a correction or source · More in Artificial Intelligence & Machine Learning