A Showcase of my Data Science and Bioinformatics Projects
This is an XGBoost project I undertook with the express intent of diving into some of the most important and useful tools in Explainable AI. I used a basic XGBoost regressor to predict used car prices on a publicly available dataset, and then I used SKLearn and SHAP to explore some of the most important tools we can leverage to gain explainable insights from the model.
A short Computer Vision project I decided to undertake to cement my knowledge in TensorFlow, using both clean and messy data.
A short walkthrough covering the main outputs of PCA with Python, but explained in lay terms for better interpretability of the technique.
This is a Random Forest prototype model developed to predict and classify a hospitalised patient’s next day’s blood test results from on its most up-to-date blood results history.
This model is a multidisciplinary collaborative approach and was created as part of the 2022 IntelliHQ x ANZICS Healthcare Datathon in Australia where it got the 1st place at the regional level (Victoria) and the 2nd place at the national level.
This is a replication and an extension of a bioinformatics project that I undertook for my PhD in Neuroscience with Monash University (Australia).
Here, I analyse the differential genetic expression of two regions of the developing cortical plate derived from bulk RNA Seq using R to better understand what drives folding of the cerebral cortex.
See doi: 10.1093/cercor/bhaa171 for the full published study.
This is a short Python project that uses Logistic Regression to build and evaluate a model that detects financial fraud using SciKitLearn and data from Kaggle, and solves the model's limitations by implementing a Support Vector Machine.
This is a short Python project that uses Multiple Linear Regression to build and evaluate a model that predicts match outcomes based on several features in data from the ATP's men league.
A short python-pptx tutorial that focuses on giving a more streamlined approach to working with this library to connect Python to PowerPoint to automate presentations.
A short Seaborn (Python) tutorial showing some more advanced plotting and labelling techniques that are usually not commonly covered in online courses or tutorials, but are quite handy for reporting purposes.