Ben Cerio

WORK HISTORY

EDUCATION

ABOUT BEN CERIO

A Ph.D. data scientist with experience in social networks, marketing, and research, I have a broad range of experience gathering, manipulating, and analyzing vast amounts of data to gain insight into a system, drive decision making, and build data products. I get excited about leveraging the structure of the data representing an ensemble to build a product that empowers the individual.Data science skillset: Supervised learning -> linear regression, boosted decision trees, random forests, k-NN, logistic regression, neural networksUnsupervised learning -> k-means, hierarchical clusteringGeneral -> SVD, natural language processing (sentiment models, topic models), time series analysis, graphsStatistical techniques -> parameter estimation, hypothesis testing, Monte Carlo methods, maximum likelihood estimationComputing expertise- Python scientific computing libraries (NumPy, SciPy, Pandas, PyPlot, scikit-learn, NetworkX)- C++(including ROOT analysis libraries, TMVA machine learning libraries, and Roostats statistical libraries)- Other: bash shell (linux scripting and command line), git, SQL, DockerLeadership- Coordinated a team of three engineers to build the front and back ends of a recommendation engine- Managed an analysis team at the largest physics laboratory in the world (CERN)- Vice President of the Graduate Student Organization in Duke\'s physics department- Head Teaching Assistant for an intro physics course at Duke, coordinating a team of four TAs- Mentored three undergraduate particle physics researchersHigh impact track record- Developed a machine-learning-based recommender system currently in production at DoubleDutch- Scientific work has been published eight papers (one cited over 650 times)- Presented results at two international conferences- ATLAS collaboration author on over 330 published papers