Get Free Synopsis PDF
We will prepare a detailed synopsis for your college submission. Includes problem statement, objectives, DFD, ER diagram, and more.
A machine learning system that scores card transactions for fraud risk in near real time and gives analysts a dashboard to review flagged transactions.
Problem Statement
Fraudulent transactions are rare and constantly changing, so simple rules either miss fraud or block genuine customers. Banks need models that handle highly imbalanced data.
Objectives
Handle class imbalance with resampling techniques
Compare logistic regression, random forest, and gradient boosting
Build a transaction scoring API
Create an analyst review dashboard
Modules
Data Analysis
Exploratory analysis and feature engineering in Python and R.
Model Training
Imbalance handling, model comparison, and threshold tuning.
Scoring Service
API that returns a fraud probability for each incoming transaction.
Analyst Dashboard
Flagged transaction queue, case notes, and model performance charts.
Expected Output
A dashboard that lists incoming transactions with fraud scores and lets an analyst approve or block the suspicious ones.
Future Scope
Graph-based fraud ring detection, streaming data pipelines, and adaptive models that retrain on analyst feedback.