After looking for some interesting data to analyze, I came across [this competition] on Kaggle. An e-commerce payment solutions company, Vesta Corporation, put together a few .csv files containing card transaction information, including the transaction amount, card type, browser of purchase, some other various basic information, and over 300 features that were engineered by the company but were left undefined in the data description. The target was contained in a column named “isFraud” and defined fraudulent transactions with a 1 and valid transactions with a 0. With the .csv files in hand, it was time to throw them into a postgreSQL database and turn this information into insights.

Before I dig in, I would like to lay out the project:

• SQL Skills (postgreSQL, SQLAlchemy, Psycopg2)
• Classification Modeling (Logistic Regression, XG Boost, Random Forest, and much more)
• Interactive Visualization (Streamlit)

#postgresql #credit card transactions #fraud