Machine Learning Resources

Distinguish between Structured and Unstructured Data

Structured Data is data that has a clear and pre-defined schema. It is usually stored in a tabular structure where the rows are the observations collected and the columns represent the features of the observations. It lends itself well to the relational database format for storage, as columns of one table can be related to those of another table using the concepts of relational database theory. An example of structured data is sales data, where the schema of a table might contain the transaction id, date of sale, product sold, and the amount. This type of data can be stored in an Excel spreadsheet or RDBMS system such as PostgreSQL. Algorithms such as linear/logistic regression and decision trees are usually suitable for structured data problems. 

Unstructured Data encompasses the wide spectrum of data that does not fall within the structured category. It can come from any source, and common examples include free-form text data, image data represented by pixels, and sound waves. Algorithms such as Neural Networks and SVM tend to work well on this type of data provided sufficient training data is available. Unstructured data is often managed through NoSQL databases such as MongoDB.

Find out all the ways
that you can