Decision Trees are a non-parametric (Meaning no underlying assumptions about the distribution of the data. So, the model is constructed based on the observed data.) supervised learning (Means label is provided to model for learning.) method used for classification (If target variable is discrete.) and regression (If target variable is conitinuous). In simplest terms this model predicts the value of a target variable by learning simple decision rules inferred from the data features. Structure of Decision Tree is like where each internal node denotes a test on an feature, each branch represents an outcome of the test, and each leaf node (terminal node) holds a class label.
Gini Impurity
Gini Impurity is a measurement of the likelihood of an incorrect classification of a new instance of data, if that new instance were randomly classified according to the distribution of class labels from the data set. So, it helps us to identify which feature is better suited to split or test (here test is the same test as used in "Decision Tree" section) the data.
Formula
If we have 𝙲 total classes and 𝑝(i) is the probability of picking a datapoint with class ii, then the Gini Impurity (𝑮) is calculated as:
On the basis of lower Weighted Gini Split, we can say that Temperature is better feature compare to Emotion for predicting whether to stay at home or not.
References
Gini Impurity Measure – a simple explanation using python
