I am a graduate student at Memorial University of Newfoundland, and this page is no longer updated; I switched to using my portfolio site to update my recent work. My Portfolio
- Python
- R
- React JS, React Native, HTML, CSS
- mySQL
- AWS EC2, S3, SageMaker, RDS, Glue
- Docker, Github Action
- This portfolio is more than just a showcase of my work; it's a reflection of my journey and skills.
- Built with ReactJS, HTML, CSS and GitHub Actions.
- This original project helps to keep track of daily to-do list. It leverages Openai API to generate inspiring messages throughout the day. Users can add new tasks, check completed tasks, remove tasks and start a new day when the day is over. An important feature is tracking the percentage of completed tasks each day throughout history.
- Built with ReactJS, HTML, CSS, Python (FastAPI) and OpenAI API.
- This article provides the development of a 3-layer Neural Network (NN) from sratch (i.e., only using Numpy) for solving the binary MNIST dataset. This project offers a practical guide to the foundational aspects of deep learning and the architecture of neural networks. It primarily concentrates on building the network from the ground up (i.e., the mathematics running underthe hood of NNs).
- Built with numpy, pandas, matplotlib, AWS SageMaker and AWS S3.
- The development of a 2-layer neural network (NN) only using NumPy. This project is a practical introduction to the fundamentals of deep learning and neural network architecture. The main focus will be on the step-by-step construction of the network, aiming to provide a clear and straightforward understanding of its underlying mechanics (i.e., the mathematics behind NNs).
- Built with numpy, pandas, matplotlib and seaborn.
- This project includes 5 repositories:
- Mitochondria instance segmentation research using Detectron2
- Mitochondria instance segmentation research using YOLOv8
- Mitochondria instance segmentation web application using Detectron2
- Mitochondria instance segmentation web application using YOLOv8
- Mitochondria instance segmentation web application using Detectron2 and YOLOv8 for better comparison
- I attempted to solve the task using 2 different tools (i.e., Detectron and YOLOv8 ). The results indicated that, for this particular task, Detectron2 demonstrated superior performance over YOLOv8. However, in some cases, YOLOv8 performed better on the task of object detection. Detectron2 was chosen to deploy on a web application for this instance segmentation tasks (visit the live demo).
- Built with Pytorch, Detectron2, YOLOv8, opencv-python, numpy, pandas, scikit-image, Flask, JavaScript, HTML, CSS, Docker and Github Action.
- This project includes 4 repositories:
- I attempted to solve the task using 4 different tools (i.e., InceptionV3, MobileNetV1, MobileNetV2 and YOLOv8). After evaluating performance metrics and processing speed, MobileNetV1 was chosen as the most suitable model for this task (visit the live demo).
- Built with Pytorch, TensorFlow, numpy, pandas, Flask, JavaScript, HTML, CSS, AWS EC2, AWS S3, Docker and Github Action.
- The inspiration for this project, along with some foundational GUI code, was drawn from TechwithTim.
- Built with pygame, BeautifulSoup, numpy, pandas.
- This R package can be used to request spatiotemporal fishing effort information from the Global Fishing Watch API and generate map plots for data visualization.
- Built with sf, dplyr, gfwr, jsonlite, magrittr, ggplot2, tigris, tidyverse, stars, raster, rayshader, INLA and stats.
-
🌱 Each day brings new growth in my journey with applied data science, especially in the realm of computer vision.
-
⚡ Fun fact: Curiosity has been my innate trait since day one.
-
📫 How to reach me: hnguyenthe@mun.ca or LinkedIn .