Second-Hand Fashion Advisory System
ML-powered web app for personalized pre-owned clothing discovery using content-based filtering and cosine similarity with adaptive Like/Dislike learning. Processed and indexed 44,000 fashion products across a full-stack architecture.
- NLP-based smart search & similar-item recommendations
- Linear Regression price prediction โ Rยฒ โ 0.72
- Deployed with Docker; 44k+ products indexed