Kendra Frederick Ph.D.
Data Scientist

Data Scientist with strong technical & communication skills. Well-versed at employing data to make business decisions.

Skills
Machine Learning
Regression
Classification
Clustering
Natural Language Processing
Dimensionality Reduction
Languages
Python
SQL
HTML / CSS
JavaScript
Libraries and Frameworks
scikit-learn
statsmodels
BeautifulSoup
Selenium
NLTK
gensim
Flask
D3
Plotly
Tableau
Data Management & Cloud Tools
PostgreSQL
MongoDB
Google BigQuery
AWS
Git/GitHub
Projects
Small Molecule Drug Predictor

  • Used CatBoost Classifier to predict whether a small molecule drug candidate would receive FDA approval based on its physiochemical properties and bioactivity.
  • Queried a 77-table SQL database to obtain data for over 1 million small molecules.

The Gangsta Score: Exploring Rap and Country Music Lyrics

  • Leveraged topic modeling with Latent Dirichlet Allocation to engineer the "Gangsta Score": the degree to which a song features Rap topics over Country topics.
  • Created a web app using Flask and D3 which returns the "Gangsta Score" for user-entered lyrics.
  • Employed other Natural Language Processing (NLP) tools such as K-Means clustering to discover cross-genre songs.

Breast Cancer Survival

  • Applied classification algorithms to predict 5-year survival in breast cancer patients using epidemiological data from the National Cancer Institute. 
  • Investigated Logistic Regression, Naive Bayes, CART, and Gradient Boosting models aiming to maximize ROC AUC. Explored over- and under-sampling to address class imbalance.

NFL Elo Rankings

  • Applied Multivariate Linear Regression to predict the change in an NFL team's Elo Ranking. 
  • Used BeautifulSoup and Selenium to scrape Football statistics from the web. 
  • Modified FiveThirtyEight's code to compare algorithms against each other.

Experience
Metis
Seattle, WA
Data Science Bootcamp
Sept. 2018 - Dec. 2018

  • Full-time, 12-week, ACCET-accredited program covering the fundamentals of data science including supervised and unsupervised learning, probability and statistics, and data visualization.
  • Project-oriented curriculum emphasized project design, data acquisition and cleaning, exploratory data analysis, statistical modeling, and communicating results.

Two Roads Brewing Company
Stratford, CT
Quality Assurance: Director, Manager
Nov. 2012 - July 2018

  • Built and managed Quality program for a start-up craft brewery. 
  • Compiled, analyzed, and interpreted data to set brand specifications and control limits. 

  • Deployed VBA/Excel-based batch file system to capture process data. Participated in design of and migration to Ignition SCADA software system.

  • Collaborated with other Departments to continuously improve processes and products. 
  • Managed annual departmental CapEx ($60-100k) and Operating ($120-150k) budgets.
  • Awarded Master Brewers Association of America scholarship to Brewing & Malting Science course.

Protein Sciences Corporation
Meriden, CT
Process Development Scientist: I, II
June 2009 - Oct. 2012

  • Developed and implemented novel characterization assays for recombinant protein vaccine. 
  • Managed project that was a major milestone in a government contract. 
  • Coordinated with external collaborators, customers, and vendors.  
  • Served as lead Scientist on a customer project billing >$100k monthly. 

Education
Yale University
Sept. 2002 - May 2009
Ph.D. Molecular Biophysics and Biochemistry
• Evaluated experimental results against thermodynamic models using non-linear curve fitting and simulation.
• Awarded Howard Hughes Medical Institute Predoctoral Fellowship.
University of Wisconsin-Madison
Sept. 1997 - May 2001
B.S. Molecular Biology