Technical Knowledge Intermediate 3 years experience
Summary
Machine learning developer with experience in supervised learning, neural networks, and healthcare analytics. Co-authored 3 IEEE publications applying ML to Long COVID prediction.
How I Apply This Skill
- Trained Random Forest classifier achieving 83.8% accuracy on Titanic survival prediction
- Built CNNs with TensorFlow/Keras achieving 84% accuracy on CIFAR-10 image classification
- Implemented Echo State Network from scratch for time series forcasting
- Used fuzzy logic techniques to handle uncertainty in symptom severity data
- Achieved AUC of 0.721 on Long COVID prediciton with custom random forest
- Applied strategic filter ordering in the Job Search Agent to reduce LLM classification costs by 99.5%, using regex-based classifiers and SHA-256 deduplication as free pre-filters before Claude Haiku 4.5 enrichment
- Fit a Holt-Winters exponential smoothing model to 188 days of global COVID-19 confirmed cases, extrapolating 30 days past the dataset end with a 95% confidence band
Key Strengths
- Model Selection: Decision Trees, Random Forest, SVMs, Naive Bayes, Neural Networks
- Feature Engineering: association rule mining, correlation analysis
- Evaluation Metrics: AUC, accuracy, precision/recall, MSE, cross-validation
- Hyperparameter Tuning: Grid search, regularization, learning rates
- Domain Application: Healthcare analytics with published research