ADS Capstone Chronicles Revised

15

Table 1 Patient Data Structure

Income, and Glucose Values. This visualization reveals there are not equivalent distributions across the numeric and ordinal variables. Such imbalances will be addressed during data preprocessing, as they could impact model training and the accuracy of predictions.

4.2.1.1 Univariate Analysis To understand the general distribution of the different data types, univariate analyses were performed using descriptive statistics for numerical and ordinal variables, and categorical counts for binary features. Figure 9 below presents the distributions of all numeric and ordinal features within histograms: body mass index (BMI), Age, Education,

217