Elewayte
Data Science intern
Developed a Hybrid Feature Optimized Model for Cyber Threat Identification focused on improving cyber threat detection accuracy and efficiency through optimized feature selection techniques. The project involved data preprocessing, feature engineering, cybersecurity dataset analysis, and implementation of intelligent detection approaches to identify malicious activities and potential threats.
Engineered a hybrid feature-optimized ML pipeline for cyber threat detection on the NSL-KDD dataset, applying RFE, Chi-Square, Mutual Information, and Feature Importance Ranking across LSTM, CNN, and XGBoost classifiers, achieving 93% detection accuracy while reducing feature dimensionality through multi-method selection.
Key skills- Machine learning
- Data Science
- Python
- Scikit-Learn
- Deep Learning
- Neural Networks
- pandas
- NumPy
- PL/SQL
- Git
- Exploratory data analysis
- Feature engineering