Overview
A travel assistance app built during a study abroad program in Europe. Uses two core ML models to help users find the right country and mover based on personal priorities like cost of living, safety, language, climate, and healthcare quality.
Cosine Similarity (CosineSimilarityModel in cos_model.py) ranks countries by computing cosine similarity on scaled user-preference features, surfacing the best matches across 30+ nations. Linear Regression (CrimeModel in lin_reg_model.py) predicts per-country crime rates for a given year using a line-of-best-fit approach, giving users a safety forecast alongside their results.
Built fullstack with Flask and Scikit-learn over 3 weeks with a team of 4, informed by conversations with EU government professionals and data experts.
Stack
Key Numbers
