Inspiration
I got inspired after reading about how people lost a lot of money from small mistakes. That made me want to create something that could help stop fraud before it even happens. So I made the AI Fraud Detector.
What it does
The AI Fraud Detector takes information from users through a simple questionnaire on a website I made with Flask. The form uses HTML, and when the user fills it out, the answers are sent to the AI model. Then the model predicts if the transaction is fraud or not.
How I built it
I started by making the form with HTML and added some CSS to make it look better and easier to use. After that, I connected it to the backend using a Python file called app.py. This file puts the user input into a pandas DataFrame to send it to the AI model. I trained the model using scikit-learn and a CSV file that I made. Then I saved the model using Joblib. After putting all the parts together, the AI Fraud Detector was ready and working.
Challenges I faced
One of the biggest challenges was a bug when I was trying to put the data into the pandas DataFrame. It took a long time to fix, and in the end, it turned out to be just a small syntax error.
Accomplishments that I am proud of
I’m really proud of the website I made. It’s simple, and user-friendly, which is great for simple projects like this.
What I learned
I learned a lot about pandas, Joblib, and AI in general. This was my first time really using some of these tools, and it made me want to keep learning more about AI.
What’s next for AI Fraud Detector
In the future, I want to add more data so the model can make better predictions. I also want it to look at spending habits and patterns to make it even smarter and more personalized.
Log in or sign up for Devpost to join the conversation.