MEGAN HOLBORN

Data Science | Data Analytics

šŸ‘‹ Hi! I'm an aspiring data scientist with a background in biochemistry and bioinformatics, who loves problem-solving and diving deep into analytics. What excites me most about data is the endless possibilities it holdsā€”every dataset tells a unique story depending on how you approach it. Iā€™m particularly interested in exploring customer preferences and behavior to retain clients and drive growth, and the analysis of clinical data to contribute to personalised medicine.

Outside of work, Iā€™m a passionate animal lover (especially cats! šŸˆ). I also enjoy long walks in nature, and exploring various creative hobbies.

PORTFOLIO

Avocado Price Optimisation

Developed an XGBoost regression model to forecast avocado demand based on price, region, and seasonality. Leveraged demand predictions to optimise pricing strategies, maximising potential revenue.
Tools: Python, Scikit-learn, Statsmodels, Pandas, Matplotlib, MLFlow
Category: Regression, Machine Learning, Optimisation
Code

IUCN Threatened Species Dashboard

PowerBI dashboard to visualise changes in the conservation status of species on the IUCN RedList from 2002 to 2022.
Tools: PowerBI
Category: Data Visualisation
PDF Download dashboard Demo

Prediction of Kindle Product Ratings

Prediction of Amazon Kindle product ratings based on sentiment using a NaĆÆve Bayes classifier.
Tools: Python, NLTK, Scikit-learn, Pandas, Matplotlib
Category: Classification, Natural Language Processing
Code Report

Database to Store Genetic Findings

A database to store genetic findings relevant to a neurodevelopmental condition called hypoxic-ischemic encephalopathy.
Collaborators: Graeme Ford
Tools: Wagtail, Django, Python, SQL, HTML, CSS
Category: Relational Databases, Application Development
App Publication

Anime Recommender

A recommender system to suggest anime based on users' historical preferences.
Collaborators: Amos Maponya, Tevin Hlebela, Refilwe Masupu, Gingirikani Mkansi, Mookodi Mokoatle
Tools: Python, Streamlit, Scikit-learn, Pandas, Matplotlib
Category: Recommender Systems, Application Development
Code App Demo

Genetic Data Analysis

Analysis of open-source genetic data to identify genetic factors in African populations that may be linked to hypoxic-ischemic encephalopathy, a neurodevelopmental condition.
Tools: Python, SciPy, Pandas, NumPy, Matplotlib
Category: Exploratory Data Analysis, Hypothesis Testing, Descriptive Statistics
Code Report