Home

Stanislava Mincheva

Data science portfolio

Stanislava Valentinova Mincheva

Final-year Data Science and Society student at Central European University with hands-on experience in Python, SQL, machine learning, and statistical modelling across healthcare and climate datasets.

View Resume Explore Projects

Current Focus

  • Entry-level data analyst and software engineering roles focused on data-driven problem solving.
  • Applied machine learning, time-series analysis, and statistical validation in real datasets.
  • Building reliable analytical workflows that translate quantitative work into clear decisions.

Highlights

CheckCells internship

Performed statistical validation of AI-based semen analysis models, supporting two FDA regulatory submissions and two peer-reviewed publications.

Healthcare modeling project

Processed 40,000+ ICU patient records from MIMIC-III with SQL and Python and built a pathogen alert classification workflow for clinical safety analysis.

Time-series forecasting

Analysed European CO2 emissions and fuel consumption data from 1990 to 2020 using regression and ARIMA models to improve out-of-sample forecasting.

Published research contribution

Co-authored work on AI-driven semen analysis for male infertility diagnostics published in the World Journal of Men's Health in 2025.

OOP In Practice

This portfolio is designed to make object-oriented thinking visible rather than keeping it hidden inside app code.

  • The website is a Quarto shell that presents projects, context, and documentation clearly.
  • The interactive components are Shiny for Python applications built around separable classes for data loading, analysis, visualization, UI composition, and orchestration.
  • The food-emissions dashboard and the MIMIC-III dashboard both use abstract loader classes and dedicated repository/analyzer/visualizer layers, showing OOP as a design decision rather than a cosmetic pattern.
  • Each app is structured so datasets, business logic, and interface logic can be extended independently.

This makes the portfolio a demonstration of both data-science outcomes and software-engineering discipline.

Contact