Rafael Carrasco
Phone: 432-***-****
Location: Irving, TX 75062
LinkedIn: rafael-carrasco
Github: erdos2n
Email: ********@******.**.***
Profile
Data Scientist with over 5 years of experience using data to structure curriculum in education. Achievement oriented professional that uses machine learning to enhance the customer experience and save on the bottom line.
Technical Skills
Programming
Machine Learning / Deep Learning
Python, TensorFlow, Keras, pyplot, bokeh, plot.ly
Web-scraping, REST API, BeautifulSoup
MatLab, Java, HTML, R
Regression, SVM, Boosting, Random Forest, Neural Networks, Natural Language Processing, Clustering, Recommender Systems, Convolutional Neural Networks, Docker
Database and Engineering
Math and Statistics
Postgres SQL, NoSQL, MySQL, MongoDB,
Hadoop, Apache Spark, PySpark,
AWS: EC2/GPU computing,
IMB Watson and DSX
Cost-benefit analysis, Hypothesis testing, Markov Chains
Multi-Arm BanditA/B testing, Cross Validation, Bootstrap sampling, Bayesian Optimization
Projects
Infinite Agency -- Using Machine Learning to Improve Ad Impressions -- https://github.com/erdos2n/Capstone
Demonstrated the efficiency of multi-arm bandit in impression modeling vs A/B testing and also applied machine learning to attribution models. After showing the benefits of multi-arm bandit, I began using weighted attribution models that highlight each ad’s contribution on the path to conversion. These insights helped campaign efficiency, effectiveness and brought the ROI up for the client.
MDP Hackathon -- Patient/Doctor Recommender -- github.com/erdos2n/Hackerthon-Athena2017
Used a K-Means cluster and recommender system to rate doctors to clients so that they may receive the best care possible.
Education
Galvanize - Data Science Immersive - June 2017
University of Notre Dame - MEd Mathematics - June 2013
Southern Methodist University - MS Computational and Applied Mathematics - May 2011
University of Dallas - BA Mathematics - May 2009
Professional Experience
Data Scientist (Consultant 1 month), Infinite Agency, Dallas, Tx
May 2017
Enhanced campaign performance using more advanced attribution models.
Used Machine Learning and NLP to streamline employee reviews with HR and Management.
Math Author, Uworld, Irving, Tx
June 2016 - February 2017
Authored original problems/explanations to help students prepare for the SAT.
Created visual representations to teach students mathematical concepts clearly.
Reviewed work from colleagues and helped them create student centered explanations.
Founding Math Faculty, Cristo Rey Dallas College Prep, Dallas, Tx
June 2015 - June 2016
Designed and constructed blended learning curriculum for Algebra I and Algebra II.
Developed an Honors by Choice curriculum using data to drive the assessments of the classroom.
Used analysis to drive curriculum and assessment that resulted in an average of 2.5 years growth per student.