Unsupervised Learning
Unsupervised LearningIt is a learning method that does not provide the corresponding category identification for the training set and is usually applicable to situations where there is a data set but no labels.
Unsupervised learning features
- The data used is unlabeled, that is, the output result corresponding to the input data is unknown, and it can only find data models and rules by itself, such as clustering and anomaly detection;
- Its purpose is to classify the original data in order to understand the internal structure of the data;
- During learning, it is unknown whether the classification result is correct, that is, no supervised enhancement is received;
- Such a network is only fed with input examples, and it automatically finds latent class rules from these examples, learns and tests them, and then applies them to new cases.
Machine learning is currently divided into supervised learning, unsupervised learning, and semi-supervised learning, and the classification criteria are whether the training samples contain human-labeled results.