Classification and Regression Tree (CART)


Classification and Regression Tree (CART)-

Classification and Regression Tree module are popularly used for alternatives of any method for regression. It is introduced by beriman at 1984. The CART follows different method for calculating the future outcomes. It is used a binary tree structure with sequential manner and that all sequence are represent a classified data. The variables are divided in tree structure and find a predicted values for future use.

The CART also used cross validation for checks accuracy. The CART model is very valuable tool for predicting Modelling and data mining. The all previous tree methodologies suffer from problem including accuracy, greediness, stability at the time of split root. The CART recover all various drawbacks about tree mining and work great.

Definition of CART– “Build’s classification or regression trees for numeric attributes means regression are categorical attributes means classification.”


The following steps are follows for in CART method-

  1. Start with root node.
  2. Split the node with more purity of data.
  3. Assigning predefined the classes to each and every node.
  4. Stop tree building when every expect of data set is visible in decision tree value check in cart.
  5. Optimal selection fallow means checks the errors in that tree.
  6. Stop tree building.

Advantages of CART-

  1. Handles data with any structure.
  2. Using machine learning in CART.
  3. the final result will summarized with logical if-then condition.

Leave a Reply