How is logistic regression different from decision trees?
Logistic regression and decision tree are both classification based algorithms but they are very different from each other
- Logistic regression looks like an equation between independent variables with respect to dependent variable but decision tree works on simple sentences which produces an outcome based on the decision statement.
- Logistic regression is a parametric model that has parameter to predict the dependent variable but decision tree is a non-parametric model in which no parameter exists.
- Assumptions can be made target variables with binomial or Bernoulli distribution but no assumptions can be made on the decision tree algorithm
- Shape of the model is predefined as logistic curve but it is not predefined in decision tree algorithm
- Logistic algorithms provide good results when independent variables are continuous in nature but decision trees produce results when the variables are categorical in nature.
- Logistic regression finds it difficult to interact with non linear variables they do not affect the performance in case of decision tree algorithm.
- Outliers and missing values reduce the performance of logistic regression but they do not affect the performance of decision tree algorithm.