Page 51 - FULL REPORT 30012024

NV has the advantage of being adaptable to different dataset types

because it can handle both categorical and numerical data. It has a

wide range of applications and can be used for both binary and
multiclass classification jobs. NV models are appropriate for huge

datasets or real-time applications since they are computationally
effective and relatively easy. NV, on the other hand, does make a

strong assumption about feature independence, presuming that, given

the class name, all features are independent of one another. In some
circumstances, this presumption might not be accurate, which would

diminish precision. NV may have trouble accurately capturing
relationships between features when there are dependencies or

correlations between them.

Despite this drawback, NV is still a useful technique, especially when

working with highly dimensional and limited information or when
feature independence is tolerable. When choosing whether NV is the

best option for a certain classification assignment, it is important to
take the unique qualities of the dataset and the underlying

dependencies among features into account.

iii. Decision Tree (DT)

Decision tree (DT) is a well-liked data mining method that can deal

with both classification and regression issues. It has many benefits,
including as interpretability, effectiveness, and adaptability. DT may

work with a variety of data types because they are adaptable and can
handle qualitative, quantitative, continuous, and discrete variables

(Hajjej et al., 2022). They are a flexible option for a variety of datasets

because they can handle both categorical and numerical data. DT
algorithms can also be used by users of various levels of skill because

they are reasonably simple to develop and interpret. Large datasets
can be handled by DT with success. But various DT algorithms can

provide various models with various degrees of accuracy. As a result,

34