Derek Castleman

Educational Data Analyst familiar with gathering, cleaning, organizing and visualizing data. Advanced understanding of statistical and analytical techniques. Highly organized, motivated and diligent with significant background in research, program coordination and education.

• Capstone project developing a machine learning model in Python predicting proportion of students living in poverty based on school district characteristics that had an RMSE of 5% at the national level and was able to get under 3% at the state level.
• Using R, extracted then analyzed data from Twitter to find out who is the most popular Stoic philosopher on the platform.
• Participated in Titanic Kaggle Contest developing a machine learning model that predicts possibility of surviving the disaster based on passenger characteristics with an accuracy of around 78% (within top 15% competing at the time).