CM@candy-severin

Graduate student exploring a career change

Master's student who has begun a career and is evaluating new directions while studying.

Work samples

A collection of work from my early career and graduate studies, showing skills and areas I am exploring as I consider a career change.

  • Agricultural Yield Prediction Pipeline
    Agricultural Yield Prediction Pipeline

    Designed and implemented a supervised ML modelling framework comparing linear regression, decision trees, and random forest to predict maize, rice, and cassava yields across Tanzania; 35 years of geospatial data (FAO, NASA EarthData, NBS), automated preprocessing, cross-validation, and hyperparameter tuning.

    Pythonscikit-learnpandasNumPyR
  • EU Socioeconomic Clustering and GDP Regression
    EU Socioeconomic Clustering and GDP Regression

    Applied K-Means clustering and linear regression across 27 EU countries using live World Bank API data; clusters mapped onto GDP tiers with clear separation between high-income Western economies and lower-income Eastern clusters, validated against regression residuals. End-to-end pipeline, code on GitHub.

    Pythonscikit-learnPlotlyWorld Bank API

About me

Master's student who has begun a career and is evaluating new directions while studying.

I am Candy Mellania Severin, a graduate student pursuing a master’s degree. I have started my career and am exploring a possible career change while I complete my studies.

Experience

Operations Officer

Gama Technologies Ltd · August 2023 - September 2025

Education

Maria Curie-Sklodowska University (UMCS) Lublin, Poland

MSc Business Analytics and Data Science · Class of 2027

Zhejiang University of Science and Technology Hangzhou, China

BSc Information and Computational Actuarial Science · Class of 2023

Skills

Graduate-level researchCareer explorationProfessional development

Agricultural Yield Prediction Pipeline

Designed and implemented a supervised ML modelling framework comparing linear regression, decision trees, and random forest to predict maize, rice, and cassava yields across Tanzania; 35 years of geospatial data (FAO, NASA EarthData, NBS), automated preprocessing, cross-validation, and hyperparameter tuning.

Pythonscikit-learnpandasNumPyR
View all work

EU Socioeconomic Clustering and GDP Regression

Applied K-Means clustering and linear regression across 27 EU countries using live World Bank API data; clusters mapped onto GDP tiers with clear separation between high-income Western economies and lower-income Eastern clusters, validated against regression residuals. End-to-end pipeline, code on GitHub.

Pythonscikit-learnPlotlyWorld Bank API
View all work