Supervised, unsupervised and semi supervised learning

Supervised learning ： The training data has both characteristics (feature) There are labels (label), Through training , Let the machine find the connection between features and labels by itself , In the face of data with only features but no labels , You can tell the label .

Unsupervised learning （unsupervised learning）： The marking information of the training sample is unknown , The goal is to reveal the inherent nature and law of data by learning unlabeled training samples , It provides the basis for further data analysis , Most of these learning tasks are studied 、 The most widely used " clustering " (clustering), Other unsupervised algorithms include ： Density estimation (densityestimation)、 Anomaly detection （anomaly detection) etc. .

Semi-supervised learning ： The training set contains both labeled sample data and unlabeled sample data , No human intervention is required , Let learners not rely on external interaction 、 Automatically use unlabeled samples to improve learning performance , It's semi supervised learning .