My work in data science began with my graduate thesis research, which involves the implementation of the unsupervised machine learning technique called the Nonnegative Matrix Factorization on many applications. Thus, there is much data analysis. Please see Thesis Abstract for more detail. This led to the publication of two papers "The nonnegative matrix factorization: regularization and complexity" and "The nonnegative matrix factorization and atomic deconvolution", co-authored by Dr. Kazufumi Ito. In addition to my graduate research, I created from scratch (with Dr. Harold Hastings) a course in Machine Learning and Neural Networks for undergraduates without mathematical knowledge beyond Linear Algebra. Please see the reflection on this course for more detail. My enthusiasm for data and its wonders has continued to grow, and thus my involvement. I have been faculty advisor on two projects in data science (please see the reflection for more detail), and I participated in a week long workshop, the SAMSI Summer School Optimization Program 2016, where there were many statistical optimization workshops. I have advised several undergraduate theses that involved data science work, and I have led undergraduate research projects in data science. I have volunteered with the organization DataKind on a project for the American Red Cross in order to gain more real world experience.
The purpose of this portfolio is to document some independent projects in the realm of data science. The documentation is meant to demonstrate my thought process and my ability to communicate my thoughts. These projects are informal reports. The first four projects are completed in RMarkdown, but Python code is available on my git.
Project in Exploratory Data Analysis
Project in Classification Part 1 - K Nearest Neighbors
Project in Classification Part 2 - Logistic Regression
Project in Clustering - K Means, K Medians, and Density-Based Spectral Clustering of Applications with Noise (DBSCAN)
Project in Topological Data Analysis - The Mapper Method
Feel free to explore my git for more projects and scripts.