Goal Label tags for products in the women clothes category
Oue team chose the following attributes to predict the labels for:
- Style
- Occasion
- Category
- Fit
- Primary Color, Subcategory Top
The approach was to clean the data and then use tokenizing, n-grams, tf-idf etc to create features. We then used neural networks for supervised learning and create prediction models.
The codes for each attribute are in their respective folders. Please run the last 4 cells of each notebook to see the predictions.