Decision tree vs SVM

A Decision tree is a tool in Machine learning that is tree like model which uses some conditions to arrive at a consequence. Each Condition is a If-Else like statement (Example When coin flipped, If heads Else Tails).

Decision tree works with limited amount of data and when features are non-monotonic. A small change in data changes the tree structure. It works will with structured data. It requires less training time. Performance is less comparatively. Due to tree like structure, it is easy to understand but difficult to scale to large dataset. It has limitation in power to learn complicated rules.

Examples where Decision tree used include:

-Predicting sales based on historical data

-Predicting defaulting loan in banking mortgages

-Choosing steps to follow in first-aid

-Choosing diet

A SVM(Support Vector Machine) is a tool in Machine Learning constructs a hyperplane to separate data into different classes in a n-dimensional space.

SVM works better with large amount of data where there is more input training data. It can also fit any data changes because of n-dimensional classification. Easy to scale to large datasets. It is powerful in learning complicated rules and efficient in performance. More training time required. But it is difficult to interpret due to complex data transfers. and to know what patters they rely on.

Examples where SVM used include:

-Classification of news articles into “business” and “Movies”

-Classification of web pages into personal home pages and others

-Validating signatures on documents

-Face detection

-Handwriting recognition

Originally published at https://www.numpyninja.com on January 30, 2021.