Aspiring Data Scientist
I'm Fiona, currently I work as an Analyst but am ready to take the next step to becoming a data Scientist. I am genuinely enthusiastic about working with data and uncovering patterns, analysing trends, and transforming raw information into meaningful insights and visualizations. I thrive on challenges and enjoy using my problem solving abilities to tackle complex data-related issues. Whether it's cleaning messy datasets or building predictive models, I am always eager to find solutions and learn something new along the way. Below are some of the key skills I bring to the table:
Here are a few of my most recent projects. I upload both completed projects and projects that I will be completing soon, so make sure to come back to see what I have completed. If you have any questions on any of my projects listed here then don't be afraid to contact me and ask.
This was a competition project from Kaggle. The goal of this project was to be able to predict the sale price of the houses by using feature engineering and working with advanced regression techniques. At the time of submission I was in the top 12%.
This is an image recognition project where I am working with neural networks. This project is about creating a model that will be able to predict oil palm plantations from satellite images.
A kaggle competition where you predict how capable each applicant is for repaying a loan. An import part of this project was how I managed the data imbalance, I handeled it by using a Pipeline and SMOTE. I also used LGBM Classifier for the model.
This time series project is to be able to predict forecasting future web traffic for approximately 145,000 Wikipedia articles.
Instead of the regular Titanic dataset that everyone knows, this dataset is about the future titanic set in Space where instead of the passengers surviving or not they are “Transported “ into another dimension. At the time of submission I was in the top 24%
This stock exchange competition is what I am currently working on. The goal of the project will compare my model against real future returns after the training phase is complete.
