1.聚类:
数据库集群是合并多个连接到单个数据库的服务器或实例的过程。有时一台服务器可能不足以管理数据量或请求数,也就是说,当需要数据集群时。SQL是用于管理数据库信息的语言。集群采用不同的形式,具体取决于数据的存储方式和资源分配方式。
2.数据库分类:
数据库管理系统可以基于几个标准进行分类,例如数据模型,用户编号和数据库分布等,如下图所示。
DBMS中分类和聚类之间的区别:
CLASSIFICATION | CLUSTERING |
---|---|
Its all about predicting the output when input data is given. | It is all about grouping data points together based on similarities among them and difference from others. |
Labeled data is provided. | Unlabeled data provided. |
This model function classifies the data into one of defined definite classes. | This function maps the data into one of the multiple clusters where the arrangement of data items is relies on the similarities between them. |
In classification data are grouped by analyzing data objects whose class label is known. | Clustering analyzes data objects without knowing class label. |
There is some prior knowledge of attributes of each classification. | There is no prior knowledge of attributes of data to form clusters. |
It is done by classifying output based on the value of input data. | It is done by grouping only the input data because output is not predefined. |
The number of class are known before classification as there is predefined output based input data. | The number of clusters are not known before clustering.These are identified after completion of clustering. |
It is considered as the supervised learning because class labels are known before. | It is considered as unsupervised learning because their is no prior knowledge of class label. |