As a Data Scientist and Software Engineer in the financial services industry, my project involved developing a machine learning model dedicated to detecting financial fraud. This encompassed data set curation, pre-processing, and model optimization.
Challenges & Solutions
The project tasks included:
- Data Set Curation for Fraud Detection: Creating a comprehensive data set specifically tailored for detecting financial fraud.
- Development of a Machine Learning Model: Building a robust model using tools like SAS Suite and Python’s data science stack.
- Data Pre-processing and Normalization: Ensuring data quality and consistency for effective model training.
- Model Performance Evaluation and Iteration: Continuously evaluating the model’s effectiveness and making iterative improvements for optimal performance.
Key technologies used were:
- Data Management: SQL, ETL, Data Pipelines
- Analytics and Machine Learning: SAS Suite (Enterprise Guide, Viya 4, Model Studio), Python Stack (Pandas, Numpy, SciKit-Learn)
- Version Control: Git
Impact and Outcome
The project led to:
- Effective Fraud Detection Model: A highly capable machine learning model that significantly enhances fraud detection capabilities.
- Improved Data Handling: Refined data curation and processing methods leading to more accurate model outcomes.
- Iterative Model Optimization: Continuous improvements in the model, ensuring it stays relevant and effective against evolving fraud patterns.
This project illustrates the critical role of machine learning in the financial sector, particularly in fraud detection, showcasing the power of data science and software engineering in tackling complex financial challenges.