Using the H&M dataset to develop different strategies for recommendation systems for different business contexts.

Outline:

1. Exploratory data analysis and visualisation – getting to know the data
2. Popularity-based recommendation system for new customers
3. NLP-based recommender system that suggests similar articles to the one that is currently in the shopping basked (before purchasing)
4. Deep learning collaborative filtering for product recommendations after purchase (“other customers who bought this also liked…”)

The original dataset can be downloaded here.