When used for a regression task such as predicting demand for a product, how does a decision tree make its predictions? For each new observation, we look at the training data and find the most similar point, and then use the corresponding target label as the prediction for the new observation For each new observation, we re-train the tree with the data including the observation, and then use the mean target value of the points at the leaf as the prediction. For each new observation, it traces the route through the tree until it reaches a leaf. It then uses the mean target value of the training points at that leaf as the prediction. We trace the route through the tree for each new observation until we reach a leaf. We then use the majority vote of the class of each point at the leaf as the prediction 1 point