 |
Welcome to my portfolio. This is an eclectic selection of data science projects that I have been working on, some of which I developed in academia, some adapted from studying machine learning, and others just out of interest. |
Neural Style-Transfer App - Link
This app was based on an assignment in the Coursera deep learning specialization. It uses a pre-trained VGG-19 convolutional neural network to merge the content of one image with the style of a second image. Gradient descent is used to optimize the pixels of the generated image.
The app was built using streamlit and allows the user to modify the hyper parameters such as the ratio of content:style, learning rate and number of training epochs.
Sweets selection - Link
Does this distribution represent an expected outcome of randomly selected sweets?
Music Database - Link
This is a project that I worked on as part of a Udacity data engineering course. It demonstrates the ETL (extract, load, transform) principle, creating a PostgreSQL database from a collection of json files. The code is written in python using the psycopg2 library which acts as a wrapper for SQL. The database schema is explained in the accompanying README file along with instructions for how to set it up on a local machine.