A method of machine learning, unsupervised learning makes computers themselves find the regularity and tendency of given data by only providing data to be learned without offering any standard for analysis or right answers. Clustering, or a method to aggregate input data into clusters using feature quantity, is used for finding the regularity and tendency.
The method, which is effective in extracting the essence and structure of data and allows computers to automatically find the correlations and patterns from huge amounts of data, is expected to be used in academic studies and data mining. Disadvantages of unsupervised learning include that it is difficult to control learning because what to learn is up to the computers, and that analyses tend to be less accurate depending on the quality of the provided data and the algorithm used for clustering.
