This project was a term assignment for the class Deep Learning. My partner and I were interested in predicting game outcomes and underdog wins in soccer and finding indicators to conclude how games will end. Our developed feedforward neural network achieved a test accuracy of around 71% for predicting the correct winner in games of the five biggest European leagues. Moreover, the analysis of features importances using SHAP values showed that the neural network decisions aligned reasonably with real phenomena as they appear in soccer for games with different scores.

This online app was a group project at the end of the class Intro to Databases. The app makes it possible to search, rank, and find related movie productions based on information taken from the IMDb database. The app uses Dash to display the front-end. In the back-end, it uses a MySQL database, which contains tables with information about the movies.

This D3 Visualization compares the development of CO2 emissions with the world's countries' GDP from 1910 to 2016. It also displays the temperature anomaly trends for all continents over the years. Hovering over the world map and line graph displays additional relevant information.

This project was my final project for the class CPSC 112 Introduction to Programming in my first year at Yale (Spring 2018). Using a modified version of backtracking, the algorithm solves Sudokus that were inputted as text file and visualizes the solution.