Page 49 - FULL REPORT 30012024

iv. Random Forest (RF)

According to Islam et al. (2021), Random Forest (RF) is an algorithm

for supervised learning. It generates a "forest" from a collection of

decision trees that are typically trained through a "bagging"
procedure. RF is one of the first and most well-known machine

learning techniques is the decision tree (DT). In order to classify data

objects into a tree-like structure, decision logics, or tests and
corresponding results, are modelled as decision trees (Uddin et al.,

2019). RF is an ensemble learning technique that combines a number
of decision trees to produce an effective predictive model. By using a

technique called bagging, each tree in the forest is trained using a
random subset of the training data, and each node also takes into

account a random sample of features. RF accomplish this by lowering

the connection between trees and boosting ensemble variety. This
enhances the model's generalisation ability and lessens overfitting.

The RF combines all individual trees' forecasts to create a final
prediction when making predictions. Figure 2.12 shows an example

of a random forest made up of three distinct decision trees. A random
subset of the training data was used to train each of those three

decision trees.

Figure 2.12 Random Forest
(Source: Uddin et al., 2019)

32