I am a developer.
As a self-taught programmer, my experience in programming comes primarily from my time in academia and Data Science bootcamp at Springboard. There, I've learned data wrangling, data visualization, data analytic techniques, and predictive modeling using Python 3 and deployed them through Flask and Heroku. I'm currently learning the basics of web design using HTML, CSS, and JavaScript to develop interactive web pages (see link to my web-development portfolio, which is still under development).
Programming Language
Scripting Language
Querying Language
Worked as a Graduate Research Assistant at Nagoya University studying chemistry for my Master's Degree. Begun to transition from "wet" lab work to computational work. Was exposed to Bash for the very first time as well as written my first bash script and working Python code towards a research project.
Worked as a Graduate Research Fellow at the University of Tennessee, Knoxville in their Energy Science & Engineering program performing chemical research in collaboration with Oak Ridge National Laboratory. It was my most active years in terms of academic research authoring/co-authoring 7 publications.
Participated and completed a 6 months Data Science Bootcamp at Springboard in preparation for joining the workforce. Studied data science concepts and data-related packages in Python and learned to query databases using SQL.
Learning to deploy machine learning models online using Flask and Heroku while learning basic HTML and CSS syntax. Created my very first web portfolio using HTML, CSS, and Javascript.
A working web application on Heroku using Random Forest to predict whether or not the content of news is fake or not. It has an accuracy of 82%.
A reposity of notebooks used to analyze, model, and predict web traffic from Wikipedia pages from July 2015 to December 2016.
Several different Python codes programed to solve molecular weights, partial pressures, and the van der Waals equation.
A object detector developed using Residual Neural Nets (ResNet) specifically for roadsigns. It works decently for roadsigns when they're large or close up (up to 95% accuracy), but poorly for smaller or more distant roadsigns (roughly 60% accuracy).
My first independent project, it's an AI-powered chatbot trained using a small subset of dialogs from the Cornell Movie Dialogues Corpus in order to generate responses. The backbone for the chatbot was LSTM (Long Short-Term Memory Neural Nets).
A guided capstone at Springboard for the optimzation of ticket prices at a mountain resort in order the get the most out of their value. The model used for optimizing ticket prices were based on the resort's prominent features in comparison to other resorts across the nation.