Implements a k-nearest neighbors classifier, tuned via grid search, to identify fraudulent transactions in a credit-card dataset.
The project reads the “Credit_Card_Fraud_Detection.csv” dataset with pandas, examines its structure, and applies StandardScaler to normalize features. It splits data into training and test sets, then uses GridSearchCV to find the best KNeighborsClassifier
hyperparameters. An error-rate vs. k plot illustrates tuning choices,and final evaluation metrics include accuracy_score, classification_report, and confusion_matrix.