What demographic features drive income level?

We compared Naïve Bayes, Logistic Regression, SVMs, Decision Trees, Random Forests, and Adaboost and agreed on Adaboost to be the most efficient model for prediction. Then we selected standardized features in Adaboost model and conducted Principle Component Analysis and K-mean clusters.

The predictability of the stock market

Use time-series, machine learning and neural network models to evaluate the predictability of the stock market.

Credit Risk prediction and Optimization

We chose LDA, Decision Tree and XGBoost to evaluate credit risk for Fulton Bank and summarized five key aspects from which Fulton Bank can optimize its credit system.