Joseph Shim

Predict Gastrointestinal symtpoms with brain data

Applied unsupervised learning methods (Hierarchical clustering, DBSCAN, and PCA) and executed the deep learning prediction model to understand the relation between the gastrointestinal symptom and the brain substrate activity using dataset > 500K participants from UK Biobank using TensorFlow in Python.

Executed analysis using High Performance Computing (HPC) machine at NYU.

Government policy predictor website

Built a website to predict COVID-19 related government policy type using Natural Language Processing and machine learning methods using over 70,000+ policy descriptions with interactive visualization using Tableau.

Used Scikit-learn and NLTK libraries (machine learning) for model building, and HTML/CSS, Flask (Python), PostgreSQL for website development (front and backend).

Measureing biasness in the US news using NLP

Applying sentiment analysis and machine learning models to detect and to predict biasness in the major US News articles.

Collaboration with Andrew Pagtakhan, Joseph Shim, Cinthia Jazmin Trejo Medina as part of final project in "Messy Data and Mahcine Learning"

Global temperature anomaly and carbon dioxide emission

"World Temperature is Increasing" which intends to emphasize the global warming effect by demonstrating an increasing trend of the global temperature anomaly as well as a constant accumulation of carbon dioxide in the earth over the last century.

Placed 2nd on Applied Statistics Student Visualization Competition.

New York University