Matr. Nr: 642365
Hover to see contact info
Matr. Nr: 640981
Hover to see contact info
Exploring the intersection of artificial intelligence and human-centered design to create meaningful solutions.
View ProjectsA collection of our human-centric AI projects demonstrating ethical AI development and practical applications.
This project creates a Django-based web application to simplify supervised machine learning tasks. Users can upload CSV datasets, explore data through visualizations, and train various machine learning models using an interactive interface. The app handles data splitting, model training, and evaluation, making it easy for users to build and assess predictive models without deep programming knowledge.
This project explores active learning methods to efficiently train a sentiment classifier on the IMDB 50k movie review dataset. Starting with a fully supervised model as a baseline, the project then simulates label-efficient learning by interactively selecting the most informative samples, aiming to achieve high accuracy with minimal labeled data.
This project focuses on building interpretable models using the Palmer Penguins dataset. It includes an interactive interface to explore decision trees and logistic regression with adjustable model complexity, as well as a counterfactual explanation tool to understand predictions and suggest minimal changes for alternate outcomes.
This project addresses the cold-start problem by letting new users rate movies while seeing how their input shapes future recommendations. Using the MovieLens dataset, it builds an interactive, user-guided system to quickly learn individual tastes.
This project trains a mouse agent in a 5×5 grid using reinforcement learning. The REINFORCE algorithm helps it find cheese while avoiding traps. Human feedback adjusts rewards to discourage eating organic cheese.