In this project I got to work with a fun dataset. The Sentinel-2 satellite images are openly and freely accessible provided in the Earth observation program Copernicus. A group of researchers put some of them together to create novel dataset consisting out of 10 classes with in total 27,000 labeled and geo-referenced images. I used a deep learning approach, utilizing the RESNET50 architecture, to find an algorithm to classify the land use and land cover of the 64x64 pixel pictures as accurate as possible.

Kiva.org is a non-profit organization that operates an online platform to facilitate micro-lending to individuals and small businesses mostly in the global south. The platform connects individual lenders with entrepreneurs and borrowers who need financial assistance to start or expand their businesses, access education, or improve their living conditions. In a recent project I got to work with some data of this organization. It was quite interesting to dive a little deeper into this field of micro credits and look at some underlying dynamics.

In this little project, I was messing around with StatsBomb soccer data. I imagined myself being part of the German Women’s national team staff, analyzing the English team right before their finals game in the European Championship 2022. After finishing my PhD analyzing the NBA, I wanted to look at the sport I played as a teenager again, to see what the data and analytics landscape looks like. So keep this in mind, when reading my thoughts and checking my plots here.

An Elo Rating, or simply Elo, is a rating system used to measure the relative skill levels of players or teams in games and sports. It was originally developed for chess by Arpad Elo but has been widely adopted in various sports, including soccer. The Elo rating system provides a numerical representation of a player’s or team’s skill, making it easier to compare and rank them. In soccer, Elo ratings can be applied to both club teams and national teams, offering a standardized way to evaluate and rank their performances.

I recently discovered this interesting data set , containing all studio album songs of one of my favorite bands of all time - Radiohead. I checked out how the band has progressed over time looking at their 9 studio albums. This meant examining hard facts like song durations and temporal profiles of their records. However I also analyzed softer dimensions like valence, lyrical density and sadness of the words sung.

This dataset is all about meteorites, where they fell and when they fell! Data comes from the Meteoritical Society by way of NASA. If you want to find out more about meteorite classifications, check Wikipedia: https://en.wikipedia.org/wiki/Meteorite_classification. I created a plot and added the code here. Link to GitHub Repository Meteorites Link to GitHub Repository

The Himalayan Database is a compilation of records for all expeditions that have climbed in the Nepal Himalaya. The database is based on the expedition archives of Elizabeth Hawley, a longtime journalist based in Kathmandu, and it is supplemented by information gathered from books, alpine journals and correspondence with Himalayan climbers. The data cover all expeditions from 1905 through Spring 2019 to more than 465 significant peaks in Nepal. Also included are expeditions to both sides of border peaks such as Everest, Cho Oyu, Makalu and Kangchenjunga as well as to some smaller border peaks.