Annual Income Prediction | Statistical Learning II: Multivariate Analysis

Predicting individuals’ annual income (“<=50K” and “>50K”) and exploring the relationship between multiple covariates and income.

Used the Census Income Data Set from the UCI Machine Learning Repository to build the models
Performed the exploratory data analysis, including observing the characteristics of each variable, drawing statistical charts of “variable vs. Income”, and observing the correlation between variables
Extracted a sample dataset, performed tunning on the Logistic Regression, CART, Random Forest, and AdaBoost models, and trained the four models using the complete dataset to predict the results
Ran the optimal models on the test data and compared multiple metrics, indicating the best performance of CART and Random Forest (with a balanced accuracy of 0.8236 )