Nghi Huynh
Entry-Level Data Scientist | Chess Strategist
Machine Learning Projects
KYMN-Mental Health Chatbot Powered by AI
Tech Stack: Python, PyTorch, Tensorflow, Scikit-learn, Matplotlib, NumPy, Pandas, ParlAI
KYMN is a smart journaling app with a chatbot to help monitor your mental health status day by day. KYMN is powered by a multi-modal ML system.
- A text-based mental health detection model using NLP techniques
- An image-based sentiment analysis using CNN models
- A pretrained NLP transformer generator chatbot model
Sartorius-cell instance segmentation using Mask R-CNN
Tech Stack: Python, Tensorflow, Matplotlib, NumPy, Pandas, imgaug
Applying Mask R-CNN to detect and delineate distinct objects of interest in biological images depicting neuronal cell types commonly used in the study of neurological disorders.
Tech Stack: Python, PyTorch, Scikit-learn, Matplotlib, NumPy, Pandas, Albumentations, OpenCV
A pre-trained ResUNet model to segment brain tumor in 2D images. The model performs instance segmentation with a mean IoU score of 90%.
Data Science Projects
Time Trends and Predictions of Mental Health and Suicide Rates based on Socioeconomic Indicators
Tech Stack: Python, tslearn, pmdarima, Scikit-learn, Matplotlib, Seaborn, NumPy, Pandas
What is the relevance of mental health for economic development in low- and middle-income countries? Why do you think we should consider mental health when discussing suicide rates? Is there any connection between mental health, suicide rates, and economics? What are the predicted mental health trends and suicide rates for the next decade? In this project, we will answer these questions by analyzing the influences of socioeconomic factors on mental health and suicide rates on a global scale, and forecasting the trends for the next decade.
COVID-19 Misinformation detection using deep learning
Tech Stack: Python, Tensorflow, Scikit-learn, Matplotlib, NumPy, Pandas
COVID-19 virus has rapidly spread around the world and affected lots of peoples’ lives. Unfortunately, the diffusion of misinformation related to COVID-19 also gets created and propagates wildly on social media and other platforms. In this project, we built a modified LSTM with one layer and two layers to detect those fake news.