Specialty Digest

Global English editionLearners’ Study Dashboard — in development
The Knowledge Library for All
Encyclopedia / Technology & Computing

Machine Learning

Specialty Knowledge entries in Machine Learning.

A fingerprint, an example of identity matching from a single reference sample

One-Shot Learning

One-shot learning is a machine learning problem in which a model must recognize a new category, or perform a new task, after seeing only a single labeled example. It is the extreme case of few-shot learning. Humans do this easily. Show a person a single picture of an unfamiliar tool and they can recognize it

One-Shot Learning Read More »

One-shot learning is a machine learning problem in which a model must recognize a new category or perform a new task after seeing just a single labeled example.

Items moving along a conveyor belt, an analogy for data arriving in a continuous stream

Online Learning

Online learning is a method in machine learning in which a model is updated continuously as each new example, or small group of examples, arrives, instead of being trained once on a complete dataset. The model improves with experience in real time. This makes online learning useful for very large data streams that cannot fit

Online Learning Read More »

Online learning in machine learning is a training method in which a model updates itself continuously as each new example or small batch of data arrives.

A forest of many trees, the picture behind the random forest ensemble of many decision trees

Random Forest

A random forest is an ensemble learning method that builds many decision trees on random samples of the data and combines their predictions, by majority vote for classification or by averaging for regression. A single decision tree is easy to understand but unstable and prone to overfitting. A forest of trees, each slightly different, cancels

Random Forest Read More »

A random forest is an ensemble machine learning method that builds many decision trees on random samples of the data and combines their predictions by voting or averaging.

A scatter of data points with a fitted trend line, the basic picture of regression

Regression (Machine Learning)

Regression is a task in supervised learning in which a model predicts a continuous numerical value, such as a house price, tomorrow’s temperature or next month’s sales, from a set of input features. It differs from classification, where the output is a category. The word has an old origin. In the 1880s the statistician Francis

Regression (Machine Learning) Read More »

Regression is a supervised machine learning task in which a model predicts a continuous numerical value, such as a price, temperature or demand, from input features.

A game of Go in progress, a domain where reinforcement learning produced champion-level play

Reinforcement Learning

Reinforcement learning is a type of machine learning in which a software agent learns to make decisions by trying actions in an environment and receiving rewards or penalties as feedback. Nobody shows the agent the correct move. It discovers, through trial and error, which sequences of actions lead to the best long-term outcome. The approach

Reinforcement Learning Read More »

Reinforcement learning is a type of machine learning in which an agent learns to make decisions by trying actions in an environment and receiving rewards or penalties.

Pieces of a jigsaw puzzle, an analogy for learning by predicting missing parts of the data

Self-Supervised Learning

Self-supervised learning is a machine learning approach in which a model creates its own training signal from unlabeled data, usually by hiding part of the input and learning to predict it. A language model, for example, reads a sentence with a word removed and tries to guess the missing word. Because the answer is already

Self-Supervised Learning Read More »

Self-supervised learning is a machine learning approach in which a model creates its own training signal from unlabeled data, for example by predicting hidden parts of the input.

A teacher with a class, an analogy for a few labeled examples guiding learning from many unlabeled ones

Semi-Supervised Learning

Semi-supervised learning is a type of machine learning that trains a model on a small amount of labeled data together with a large amount of unlabeled data. It sits between supervised learning, which needs a label for every example, and unsupervised learning, which uses none. The approach is useful because labeling is often slow and

Semi-Supervised Learning Read More »

Semi-supervised learning is a type of machine learning that trains a model on a small amount of labeled data together with a large amount of unlabeled data.

Examples of handwritten digits from the MNIST dataset, a classic labeled dataset used in supervised learning

Supervised Learning

Supervised learning is a type of machine learning in which a model learns from examples that come with the correct answers, called labels, so that it can predict the right output for new, unseen inputs. The word supervised refers to the labels, which act like a teacher marking each example. A model might be shown

Supervised Learning Read More »

Supervised learning is a type of machine learning in which a model learns from labeled examples to predict the correct output for new inputs.

A line dividing two sides of a road, an analogy for the wide-margin boundary an SVM seeks

Support Vector Machine

A support vector machine, or SVM, is a supervised machine learning algorithm that classifies data by finding the boundary that separates the classes with the widest possible margin. Among all the lines that could divide two groups of points, the SVM chooses the one that stays as far as possible from the nearest points on

Support Vector Machine Read More »

A support vector machine is a supervised machine learning algorithm that classifies data by finding the boundary that separates classes with the widest possible margin.

Runners passing a baton in a relay race, an analogy for passing learned knowledge from one model to another

Transfer Learning

Transfer learning is a machine learning technique in which a model trained on one task is reused as the starting point for a different but related task. Instead of training a new model from scratch, which needs large amounts of data and computing time, a practitioner takes a model that has already learned useful patterns

Transfer Learning Read More »

Transfer learning is a machine learning technique in which a model trained on one task is reused as the starting point for a different but related task.