Providing TA services after completing a 28-week immersive data science bootcamp emphasizing practical skills in machine learning, data engineering and analytics, natural language processing, and business communications. Notable projects from the bootcamp include:

Web App To Discover Music Through Genre Adjacency 

• Used Spotify's API to create a MongoDB database mapping musicians to their performed musical genres.
• Developed an algorithm to establish genre profiles for musicians and determine cosine similarity to user input with Sklearn.
• Deployed a Streamlit web app to facilitate end-user interaction as an alternative means for music discovery.

Captioning Images For Blind Users

• Created a deep learning model using a convoluted neural network to apply computer vision to >30,000 images taken by blind users.
• Paired the model from above with a second deep learning model leveraging a recursive neural network with NLP techniques to generate captions based on image content, achieving a contextual accuracy deemed useful by a small sample of end-user tests.

Classifying Mushroom Toxicity

• Created a classification model to determine the edibility of common mushrooms of the Agaricus genus.
• Used ensembling techniques with Random Forest, Decision Trees, k-Nearest Neighbors, and XGBoost to build predictive models.
• Investigated SHAP values to assist in feature selection and optimize hyperparameters, boosting model accuracy to 96%.
  

Pricing Campervans With Regression Analysis

• Collected price and feature data from Vanlifetrader.com using a custom webscraping script featuring BeautifulSoup and Selenium.
• Applied rigorous regularization (Ridge, LASSO) techniques alongside detailed feature engineering to create a model with >70% accuracy.
• Identified high-value features of campervans for the purposes of effectively marketing a fairly priced product as a private seller.

• Reviewed and updated spatial data for various parts of the globe for the NGA's Janus Geography project using ESRI ArcMap.
• Performed quality control and quality assurance for outgoing cartographic projects.
• Assisted in the development of proprietary machine learning tools for in-house GIS applications. 

• Conducted multiday safaris of various US bioregions (Mountain West and Desert Southwest).
• Created custom, educational narratives of landscapes, including complex ecological interactions, biology, geology, natural history, and culture.
• Managed small groups of clients with widely ranging demographics, interests, abilities, and perspectives.

• Investigated western wildfire regimes and biodiversity with extensive timeseries and landcover analyses.
• Provided technical GIS support for non-technical university clients.