A Decision Tree is a way of displaying an algorithm containing only conditional control statements. It uses a tree-like structure for decisions and their possible consequences including chance events. A Decision tree consists of Decision nodes represented in square type, Chance nodes typically represented by the circle, and Endnotes represented in triangles. They are most commonly used in operations research and operations management. We can also descriptively use the decision tree for calculating conditional probabilities. The decision tree algorithm fits in the category of supervised learning with the help of the algorithm we can solve regression and classification problems. The structure of the algorithm is of tree type in which each leaf node corresponds to a class label and the internal node of the tree represents the attributes. The discrete attributes are used in the decision tree for representing any Boolean function. The decision tree is simple to understand and interpret; it requires little data preparation but the cost of using the tree is logarithmic in the context of data points used for training the tree. It can handle both numerical and categorical data. It also performs well when assumptions are violated by the true model from where the data was generated. Check out our PG Course in Machine learning Today.

Read More