Description

Before starting, I will explain the 3 classification models that I use:

KNN = The k-nearest neighbors algorithm, also known as KNN or k-NN, is a non-parametric, supervised learning classifier, which uses proximity to make classifications or predictions about the grouping of an individual data point. While it can be used for either regression or classification problems, it is typically used as a classification algorithm, working off the assumption that similar points can be found near one another.

Decision Tree = A decision tree is a decision support tool that uses a tree-like model of decisions and their possible consequences, including chance event outcomes, resource costs, and utility. It is one way to display an algorithm that only contains conditional control statements. Decision trees are commonly used in operations research, specifically in decision analysis, to help identify a strategy most likely to reach a goal, but are also a popular tool in machine learning.

Random Forest = Random forest classifier is a classification method consisting of a collection of decision trees supported by training data and independent random features with different features. Random forest is an algorithm that is able to classify large amounts of data accurately and the end result is obtained from determining the root node and ending with several leaf nodes.

Now, let's start from the initial stage:

Importing the libraries

Loading or read the datasets

Exoloratory Data Analysis

there are some statistical information about the dataset.

data correlation.

Data preprocessing

Handling missing value.

Filling the missing.

Handling the outliers.

Note : the gender column as was mentioned before has 3 categories: Female, Male, Other. By looking at the other category we will find that has only one record, so we can drop this record.

Encoding

Gender column encoding.

Scaling

As notice the range of columns like bmi,avg_glucose_level and age differes from the range of columns like Residence_type, work_type, etc.. So to avoid that one feature being demonstrated by the others, we need to do feature scaling to make all the features almost have the same range.

Feature selection

From the correlation table, we can just keep the features which are highly correlated to each other. We are going to keep age, hypertension, heart_disease, avg_glucose_level and bmi.

Balanced the data

Split Data

Modelling

KNN

Decision Tree

Random Forest

Conclusion

Random forest is the best model for this data