Nuclear Atypia Scoring of Breast Cancer Histological Images Using Transfer Learning

Developed a convolutional neural network (CNN) for classifying breast cancer histological images to three classes based on nuclear atypia. The project utilized transfer learning by using the Inception-v3 CNN with pretrained weights. The final network achieved 0.72 accuracy score on the testing set.

Machine Learning to Aid Diagnosis of Inflammatory Bowel Disease Using Metagenomic Data

In this project, I compared the ability of three classifiers to predict the presence of inflammatory bowel disease based on metagenomic data collected from human gut samples. The data consisted of relative abundances of microbes in the gut and included samples from healthy people and IBD patients.

Submillimeter-wave Holograms and Imaging Neural Network

As a research assistant at the Department of Electronics and Nanoengineering I participated in a project that developed a submillimeter-wave imaging application with a deep learning backend. I was tasked with testing the neural network-based image reconstruction by varying the size of the input data

Clustering of OECD Countries Based on Annual Mortality Rates for Different Disease Groups in 2020

In this project, I used agglomerative hierarchical clustering to group countries based on their annual mortality rates for nine disease groups in 2020. The dataset included 31 countries of which 26 were OECD and 5 non-OECD countries. The code was implemented in R.