Projects
Explore some of the projects I’ve worked on, showcasing my expertise in data science, machine learning, and AI solutions development.
Heart Disease Prediction
This project predicts heart disease risk using various machine learning techniques, including Logistic Regression, Decision Tree, Random Forest, and Support Vector Machine models. It emphasizes my skills in data preprocessing, feature selection, and hyperparameter optimization using GridSearchCV.
View WorkHate Speech Detection
This project addresses the critical need for automated hate speech detection, aiming to create safer and more inclusive online spaces. Using advanced machine learning techniques, I fine-tuned the BERT model specifically for hate speech classification. Through this approach, I utilized Optuna for hyperparameter optimization, which allowed me to enhance model accuracy and efficiency. I also froze the lower encoding layers of BERT, preserving its foundational understanding of natural language acquired during pre-training, while fine-tuning higher layers for the specific task of hate speech identification. This project highlights my expertise in model tuning, hyperparameter optimization, and leveraging transformer models for real-world applications.
View WorkAI Ethics: The Graham AI Ethical Framework
This essay presents a comprehensive framework for ethical AI in high-stakes fields, applying five core principles—Privacy, Transparency, Fairness, Accountability, and Autonomy—to the use of AI tools in financial trading for pension funds. The framework provides actionable guidelines for responsibly managing sensitive data, ensuring transparency in AI-driven decisions, promoting fairness, and maintaining human oversight to safeguard public trust and regulatory compliance.
Read Essay