The following is a collection of some projects, I've conducted so far.
Click here to get back to the previous page.
- Medium Blog: From time to time I write Medium blog posts here on data science related topics, which have gained quite some attention. I do not get to do it as much as I would like to anymore, but hope to fill it again soon with quantum-related blog posts.
- Interdisciplinary Project on Brain Metastases: I worked on an interdisciplinary project with physicians on Radiomic Profiling of Brain Metastases. In particular, we tried to improve existing survival models, by including radiomic data. Here is the link to the poster we created at the end of the project.
- Training and Fine-Tuning LLMs: Among the most fascinating topics in Deep Learning I encountered, is Natural Language Processing. Because I wanted to understand how such models are trained, I trained my own DistilBERT model and fine-tuned it for Question Answering. The code is open-sourced here. Furthermore, I wrote a blog post on pre-training here and on fine-tuning here.
- Predicting Income Levels of Countries: Furthermore, as one of my first projects, we conducted a Machine Learning project on predicting the income levels of countries. It includes data collection, cleaning, model training and evaluation, and was an extremely valuable learning experience for me, as it made me think a lot about data collection and biases in data. The code is open-sourced here and you can find the related blog post here.
- Automated Machine Learning: Also, I really enjoy algorithms and read a lot about Automated Machine Learning at some point. I thought the Hyperband algorithm was quite interesting, and I tried implementing it. The code is available here, a blog post on AutoML here and the description of the implementation here.