I Built a BFIU-Compliant AML Detection System in Python (Here's Why the Kaggle Approach Doesn't Work)
Most AML tutorials end with a confusion matrix and a 99% accuracy score. Here's why that doesn't work — and what I built instead. I've been working in fintech compliance data for a while. The one thing I kept noticing: every "fraud detection project" on GitHub or Kaggle uses the same dataset — the UCI credit card fraud dataset from 2013. It has 284,000 rows, 30 features labeled V1-V28, and approximately zero explanatory value for anyone who wants to understand how financial crime actually works. So I built something different. The problem with the standard approach Real transaction monitoring engines don't work like Kaggle competitions. They don't take a CSV, train a model, and output a probability score. They work like this: A rule engine runs first — deterministic, auditable, regulatory-cited rules that generate alerts Those alerts get scored and triaged by risk tier An ML layer reduces false positives among the high-risk alerts ...
This portfolio presents a strong combination of practical data science, machine learning, visualization, and business-oriented experience. The emergency severity prediction project is particularly interesting because it applies XGBoost to more than 80,000 patient records, while the dashboard and data-cleaning projects demonstrate the ability to work with both analytical models and real-world datasets.
ReplyDeleteThe data cleaning work using Pandas, including handling null values, duplicates, type conversions, and outliers, highlights an important part of any reliable analytics workflow. A solid Data Analysis Course can help strengthen these skills and provide a structured understanding of how raw datasets are prepared for meaningful analysis.
The interactive sales dashboard is another valuable addition because it demonstrates how analytical results can be communicated through revenue trends, regional breakdowns, and KPI visualizations. Learning through a Data Visualization Course can further develop the ability to transform processed data into clear and useful visual insights for business stakeholders.
ReplyDeleteThe XGBoost-based ESI prediction project also demonstrates practical experience with classification, feature preparation, and model development using Python and Scikit-learn. For students seeking hands-on exposure to predictive modeling and machine learning applications, Machine Learning Projects for Final Year can be a useful way to explore similar project-oriented challenges.
ReplyDelete