Hypertension Clinical Trial Dropout Prediction
Predicted dropout and adverse event risk in Phase III clinical trial data (~1,000 patients across 50+ sites) using recall-optimized Gradient Boosting classifiers.
My journey began in science, where I spent over four years as a process chemist designing experiments and developing routes to deliver pharmaceutical drugs for clinical trials. Along the way, I discovered the power of data science to uncover patterns and drive better decisions, which led me to complete certifications from the University of Cambridge and Google.
Since then, I’ve applied machine learning, deep learning, and time series forecasting to projects spanning clinical trials, patient data, and a forecasting challenge with the Bank of England. By combining scientific rigour with domain expertise in chemistry and pharmaceuticals, I focus on building interpretable, data-driven solutions that deliver real impact in healthcare and life sciences.
At the heart of my work is curiosity. I believe the quality of insights is limited only by the quality of the questions you ask and it’s this mindset that drives me to explore data with rigour and insight.
Here are a few Python libraries I’ve been working with recently:
pandas, NumPy, scikit-learn, SHAP
TensorFlow, Keras, PyTorch
spaCy, NLTK, HuggingFace transformers
statsmodels, pmdarima, sktime, LightGBM, XGBoost
matplotlib, seaborn
Predicted dropout and adverse event risk in Phase III clinical trial data (~1,000 patients across 50+ sites) using recall-optimized Gradient Boosting classifiers.
Minimised missed-stroke events on a severely imbalanced dataset by training a recall-focused Neural Network classifier.
Enabled proactive inventory planning by forecasting weekly and monthly Book demand.
Studied a Masters Level Qualification' at University of Cambridge.
Studied a Master of Chemistry at University of York.
My inbox is always open. Let’s connect about data science opportunities in healthcare and beyond!