1- Data description and Research question
In this case study, the Framingham heart study cohort data set is used to train and test the models for heart disease classification. The data is collected by examining the town residents from Framingham, Massachusetts, from an ongoing cardiovascular study (Kaggle,2021). This dataset contains over 4,000 records and 16 columns. This dataset was collected initially to study the prevalence of several cardiovascular diseases and the risk factors associated with them. This dataset helps us explore the patterns of CVD and the risk factors over time that affect our lives. The Framingham study was started in 1948 under the supervision of the U.S. Public health Service, which was then transferred under the control of the new National Heart Institute. Participants, including men and women from the town of Framingham, were sampled, and studied to identify the concept of risk factors and the joint effects of CVD, which also facilitates specialized studies.
Table 1- Description of the features
Feature name Data type Description
male Integer Represents the gender of the participant
age Integer Represents the age of the participant
education float Represents the education level of participant
currentSmoker Integer If the participant is a current smoker or not
cigsPerDay float Number of cigarettes the participant smokes per day on an average
BPMeds float Checks if the patient is on blood pressure medication
prevalentStroke Integer Checks if participant had a stroke previously
prevalentHyp Integer Checks if patient is hypertensive
diabetes Integer Checks if participant has diabetes
totChol float The total cholesterol level of participant
sysBP float Checks the systolic blood pressure of the patient
diaBP float Checks the diastolic blood pressure of participant
BMI float Checks body mass index of the participant
heartRate float Check the heart rate of the patient
glucose float Checks the glucose level in the body of patient
TenYearCHD Integer Checks for the ten-year risk of coronary heart disease in the patient
Based on the problem statement, I have framed a research question that I am going to answer with the help of the is project is given down below.
To identify the most relevant/risk factors of heart disease as well as predict the overall risk of whether the patient has a 10-year risk of future coronary heart disease (CHD).
2- Data preparation and cleaning
In this stage, the data preparation is done, which includes data preprocessing. Some of the steps that we will follow in data preprocessing are given below.
• Checking for the column with wrong data types assigned and fixing typos errors.
• Checking for duplicate values in the dataset. To identify duplicate values duplicated function is used, and then we remove it.
• Checking for the missing values in the dataset and imputations on missing values.
• Check for anomalies in the dataset and remove them. To identify the anomalies, a boxplot is used, and then the Interquartile range's value is used to remove them (Ngare and Kennedy, 2019).
With the help of the glimpse function, I can observe that categorical variables are given in integer type. To find out the duplicate values duplicated function is used; the dataset contains zero duplicated values. To find out the missing values, the data explorer library is used. Several columns contain missing values, such as glucose, education, bpMeds, totChol, cigsPerDay, BMI, and heart rate. The mice library was used for computing the missing values.
To find out the outliers box plot was made. From the below figure, you can see the columns like BMI, glucose, totChol, cigsPerDay, diaBP, heart rate, and sysBP. Most of the columns were having outliers.
Figure 1- Boxplot of numerical column
After that, the outlier values were stored in a result variable, and then we removed them from the dataset. After the data preprocessing was done, then I moved to the data analysis section.
3- Exploratory data analysis
In this section, EDA was conducted to uncover the hidden pattern and extract important features from the dataset. The data analysis helps improve the analytical process's transparency (Ho Yu, C, 2010).
In this project, some of the steps taken in doing EDA are given below.
1. Performing univariate and bivariate data analysis on the numerical and categorical variables with the help of scatter plots, density plots, bar charts, and many other graphs were used.
2. A correlation plot was used to check the relationship between the variables.
3. The Shapiro test is used to check the normality of data distribution.
4. I am using the unsupervised learning technique like principal component analysis for doing the exploratory data analysis.
3.1- Univariate data analysis
For Numerical features
The histogram was made for univariate analysis to analyze the numerical columns. The graphs of the numerical column are shown down below.
DescriptionIn this final assignment, the students will demonstrate their ability to apply two ma
Path finding involves finding a path from A to B. Typically we want the path to have certain properties,such as being the shortest or to avoid going t
Develop a program to emulate a purchase transaction at a retail store. Thisprogram will have two classes, a LineItem class and a Transaction class. Th
1 Project 1 Introduction - the SeaPort Project series For this set of projects for the course, we wish to simulate some of the aspects of a number of
1 Project 2 Introduction - the SeaPort Project series For this set of projects for the course, we wish to simulate some of the aspects of a number of
| Sun | Mon | Tue | Wed | Thu | Fri | Sat |
|---|---|---|---|---|---|---|
23 | 24 | 25 | 26 | 27 | 28 | 1 |
2 | 3 | 4 | 5 | 6 | 7 | 8 |
9 | 10 | 11 | 12 | 13 | 14 | 15 |
16 | 17 | 18 | 19 | 20 | 21 | 22 |
23 | 24 | 25 | 26 | 27 | 28 | 29 |
30 | 31 | 1 | 2 | 3 | 4 | 5 |
