A machine learning approach where a model learns from labeled examples, each pairing an input with the correct output.
In supervised learning, the training data includes answers. To build an email spam filter, you provide thousands of messages already labeled as spam or not spam. The algorithm learns patterns linking features of each message to its label, then applies those patterns to new messages.
Supervised tasks fall into two broad types: classification, which predicts categories such as fraud or not fraud, and regression, which predicts numbers such as house prices or delivery times. Common algorithms include logistic regression, decision trees, random forests, gradient boosting, and neural networks. The main challenge is obtaining enough high-quality labeled data, which can be expensive and time-consuming. Poor or biased labels lead directly to poor or biased models.
A branch of AI in which computers learn patterns from data to make predictions or decisions without being explicitly programmed for each case.
RegressionA statistical and machine learning method for predicting a continuous numerical value based on the relationship between variables.
Unsupervised LearningA machine learning approach that finds patterns and structure in data that has no labels or predefined answers.
A plain-English guide to reinforcement learning: agents, environments, rewards and policies, with game, robotics and chatbot examples and where it fits in business.