SmartWatch Health Data Analysis

Project Overview

This project performs Bivariate and Multivariate Analysis on a health dataset to explore relationships between categorical and numerical variables. Various visualization techniques are used to understand the impact of factors like heart rate , blood oxygen level , steps taken in a day , activity level and sleep duration on stress level of an individual. Part of the assignment given in the course DAI-101.

Dataset

The dataset being used is the health data statistics of 10000 randomly selected smartwatches . Contains information about the used factors written above

Features Analyzed

Categorical Variables

• Sleep Duration (hours)

• Activity Level

• Stress Level

Numerical Variables

• User ID

• Heart Rate (BPM)

• Blood Oxygen level (%)

• Step Count

Exploratory Data Analysis (EDA)

Univariate Analysis

• Data Description: Mean, Median, Mode, and Interquartile ranges of all the numerical variables.

• Skewness: Skewness of all the numerical variables used.

• Frequency Distribution: Frequency charts representing the frequency distribution of all the categorical variables.

• Histograms: Histograms representing all the numerical variables is used.

• BoxPlots : Boxplots of all the numerical variables

BoxPlots and Frequency charts are also present when the data was not cleaned

Bivariate Analysis

• Heatmap : Between every pair of continuous variables

• Box , Violin and Bar plots : For comparing all the numerical variables with all the categorical variables

• Scatter Plots: Understand relationships between Heart Rate and Sleep Duration (Based on steps taken ).

Multivariate Analysis

• Pair Plots: Visualize multiple variable relationships simultaneously.

• Heatmap: Analyze correlation among numerical features all at once.