I'm Olivera, a data scientist with an interdisciplinary background in physics, cognitive science, and machine learning. I specialize in Bayesian (predictive) modeling, forecasting, causal inference, and data visualizations.
About Me
I’m a data scientist with nine years of experience in Bayesian inference, predictive modeling, forecasting, and causal inference. I’ve used these methods in fields such as market research, public health, and environmental sciences. I’m also interested in the interpretability of machine learning methods and data visualizations, which help make complex data easier to understand and use.
I’m currently based in San Diego, CA. I volunteered and worked with Data Science Alliance, a tech non-profit, on a model to predict food demand for food banks. Before that, I worked at Michael Bauer Research in Nuremberg, where I focused on predictive modeling, econometrics, market research, and causal analysis.
I have a PhD in Cognitive Science from Osnabrück University, where I worked on interpretable machine learning for real-world applications. Previously, I studied Medical Physics (BSc) and Electrical and Computer Engineering (MSc) at the University of Novi Sad, Serbia.
In my free time, I like hiking, making analog photos, or creating collages from old tickets, newspapers, or similar materials.
Work Experience
2024: Data Science Alliance - Data Scientist & Volunteer
Worked on predictive models for food demand and homelessness in San Diego County, using CDC, US Census, and HUD data.
Applied the Responsible Data Science framework developed by Data Science Alliance and integrated state-of-the-art research into modeling.
2019 - 2023: Michael Bauer Research GmbH - Data Scientist
Worked on geospatial models to predict daytime population shifts across Europe and North America, incorporating consumer behavior, market trends, and socio-economic data.
Contributed to developing a geospatial model for forecasting unemployment rates across Europe and North America.
Worked on predictions of economic indicators such as purchasing power, consumer styles, and income.
Defined and validated KPIs for economic forecasting.
Built tools and dashboards to visualize complex data and automated processes to improve efficiency.
2015 - 2019: Institute of Cognitive Science - Research Assistant
Researched the interpretability of machine learning models, focusing on their application in high-stakes decision-making.
Developed three machine learning models for different applications: predicting epileptic seizures, forecasting infectious diseases, and predicting environmental variables.
Won a special prize at the 'hack4health' hackathon and interned at the Robert Koch Institute in Berlin. There, I led a project to develop a Bayesian model for predicting the spread of infectious diseases.
Collaborated with the Juelich Research Center to develop a Bayesian hierarchical model for predicting environmental variables.
Taught a semester-long course on ensemble methods in machine learning and mentored bachelor and master theses.
Education
2015 - 2023: PhD studies in Cognitive Science, Osnabrück University, Germany
2013 - 2014: Master studies in Electrical and Computer Engineering, University of Novi Sad, Serbia
2009 - 2013: Bachelor studies in Medical Physics, University of Novi Sad, Serbia
Volunteering
2024 - present:
Women in Tech San Diego - Leadership Team Member
Women Coders Inclusive Org San Diego - Volunteer
2024: Data Science Alliance - Volunteer
Tech Stack
Areas of Expertise: Data Science, Machine Learning, Data Visualization, KPI Framework Development
Data Science Methods: Predictive Modeling, Time-series Analysis, Forecasting, Bayesian inference, Probabilistic Modeling, Generalized Linear Models (GLM), Markov Chain Monte Carlo, Causal inference, Spatio-temporal models
Programming Languages: Python, SQL, R, Alteryx
Programming Libraries: Jupyter, Pytorch, scikit-learn, SciPy, Pandas, OpenCV, NumPy, PyMC, Shapely, tidyverse
Data Visualization: matplotlib, ggplot2, Plotly, seaborn