Two machine learning models: one built with a Convolutional Neural Network and the other implementing K-Nearest Neighbor were designed to detect fraudulent Credit Card Transactions.
This project was made as a final requirement for a course on machine learning. It was run on google colab and the models were trained using this dataset from Kaggle
A python notebook with all the details separated into sections is provided as well as .csv files with the datasets fit for both models.
Once all the files are downloaded, place them in a google drive folder while taking note of the location of the folder in your google drive.
Open the python notebook in a google colab, then edit the file path in the data set up portion of the code. Here's an example.
#Read data from CSV Files
from google.colab import drive
drive.mount('/content/drive')
data_3 = pd.read_csv('drive/MyDrive/creditcard_3.csv')
data_28 = pd.read_csv('drive/MyDrive/creditcard_28.csv')
data_demo = pd.read_csv('drive/MyDrive/demo_set.csv')
knn_data_demo = pd.read_csv('drive/MyDrive/creditcard.csv')Run each section of the code. Both models would display their respective confusion matrices by the end of their demos.