Banu Atav

Skills

• Python

Data Scientist

Graduate in Econometrics (MSc.) with over 4 years Data Science experience and about 1.5 years industry experience in building data-driven automation solutions using machine-learning. Completed +/- 10 projects that range from proof-of-concept/ideation phases to deployment of the solution. Fluent in Python and experienced in R. Experience with Azure & GCP AI-services.

Education

• MSc, Econometrics and Management Science

  Erasmus University Rotterdam, The Netherlands • Honors: graduated cum laude with GPA 8.7/10.0 (top 5%). GPA translates to 4.0/4.0 in US. • Relevant Coursework: Multivariate Statistics and Data Visualization, Applied Micro-econometrics, Bayesian Econometrics, Advanced Time Series Analysis and Asset Pricing. • Thesis: Non-parametric Bayesian Forecasts Of Election Outcomes. Can be retrieved from https://thesis.eur.nl/pub/44118 Elaborates on the necessity of flexible models in election forecasting by modeling individual-level, discrete data on voting intentions using a Dirichlet Process mixture model. Inference is done using a hybrid MCMC scheme that incorporates both Gibbs and Metropolis-Hastings algorithms. Code is made publicly available in the form of an R package on: https://github.com/banuatav/DPMMmlogit, which exploits C++ code to greatly improve on the speed of the sampler.