Brief About Machine Learning Techniques

The research discussed some machine learning techniques used to measure rainfall power. These techniques depends on how appropriate the features are. That is why there is different levels for each technique which will lead to making them better by time. The Five techniques are illustrated in the figure bellow.

C4.5 Algorithm

It is one of the best tree classification algorithms used in measuring rainfall prediction. Its pseudo code can be noticed from the research in table 1 as below.

Naïve Bayes

This technique classifies the data using probability. Take in consideration that assumptions between the features are independent. The pseudocode used to calculate the rainfall power in this method is mentioned here in table 2.

Support Vector Machine

This method simply divides the data into X and Y hyperplanes. Each class label is addressed alone. In other words, class X and not class X are classified alone, and class Y and not class Y are classified alone. The pseudocode can be described as below in table 3.

Neural Network

This method tries to use Artificial Intelligence in Neural Networks trying to make the computer simulate the way of how humans brain think. Each neural unit is connected to other similar neural units, just like how the nervous system work in human beings. These systems can learn by themselves and develop by time, so they are not programmed like other systems. Table 5 elaborates the algorithm used and how this Neural Network works.

Random Forest

.This technique can be used in many tasks, one of them is prediction. It is used worldwide because it is simple and can be used diversely. Table 5 illustrates one example of using Random Forest algorithm.

All the information above are explained in details in the research attached to the link below:

https://www.researchgate.net/directory/publications