This dataset contains files that represent transformed/computed datasets for the H&M competition. Made public for the following notebook: https://www.kaggle.com/code/vincentbrunner/h-m-image-embeddings. Files:

• combined_article_embedings: see notebook
• article_embedings: see notebook
• 10k simple triplets: see notebook
• sample candidate: see notebook
• aggregated_articles_per_month: needed for private notebook(aggregated, scaled buys per month per customer)
• mean_unique_features_over_months: needed for private notebook(attempt to measure the change in buying behaviour over time) -> the last 2 files come from a private customer segmentation notebook and were stored here cause I wanted all my H&M helper files in the same place