ADS Capstone Chronicles Revised
8
Figure 2 Bar Plot of Type of Service and Average Total Payment by Type of Service
Figure 3) were used during multivariate nongraphical analysis to identify the relationship between numeric variables before and after the missing values were handled. This step is crucial for identifying multicollinearity, where certain features may be highly correlated. High correlation among features can lead to redundancy and may affect the performance of the analytical model.
When performing univariate nongraphical analysis, a summary of unique values for categorical variables was developed to provide insight into the distribution and cardinality of the variables. Histograms were created to visualize the key numeric variables when performing univariate graphical analysis to observe dispersion and central tendency. Correlation matrices (see
132
Made with FlippingBook - Online Brochure Maker