Summary#

Experience building custom neural network architectures and applying machine learning to real-world problems. Emphasis on implementing algorithms from scratch to understand the mathematics, not just using libraries. IEEE-published researcher applying ML to healthcare data.

How I Apply This Skill#

• Implemented Echo State Network from scratch using Python and NumPy for k-step ahead forecasting
• Built custom hyperparameter optimization pipelines with cross-validation
• Applied data mining and classification techniques to healthcare data
• Published 3 IEEE papers on predicting Long COVID cases using ML
• Mined association rules achieving >0.8 confidence scores
• Implemented genetic algorithms for optimization problems

Key Strengths#

• From-Scratch Implementation: Neural networks using only NumPy
• Time Series: Echo State Networks, k-step ahead prediction
• Data Mining: Association rules, Apriori algorithm, feature selection
• Model Evaluation: Cross-validation, MSE analysis, hyperparameter tuning
• Research: IEEE publications demonstrating rigorous methodology

Related Projects#

• Echo State Network - Custom neural network from scratch
• Long COVID Prediction - 3 IEEE publications
• Line of Best Fit - Genetic algorithm optimization
• Country Data Analysis - PCA and clustering