Supervised learning

Supervised learning trains a model on a labeled dataset (input-output pairs) to map features to a target output: It predicts outcomes based on known examples.

Supervised learning is a core machine learning paradigm: The algorithm learns a function by analyzing a large, labeled training dataset (e.g., 10,000 images with 'cat' or 'dog' tags). This process enables accurate prediction on new, unseen data. Key applications divide into two types: Classification predicts a discrete label (e.g., using Logistic Regression to flag an email as 'spam' or 'not spam'). Regression predicts a continuous value (e.g., using Linear Regression to forecast a house price based on 5 specific features). Algorithms like Support Vector Machines (SVM) and Random Forest are standard tools in this high-accuracy, pattern-recognition approach.

https://www.ibm.com/topics/supervised-learning