Problem and data domain
In this chapter, we will employ MLP models to understand the relationship between customers’ preferences and their purchasing patterns using the Fashion-MNIST dataset. MLP models offer a powerful framework for image classification tasks, allowing us to predict the type of clothing or accessory a customer is likely to purchase based on their interactions with the online store. By uncovering patterns in customer preferences, the e-commerce store can personalize recommendations and optimize inventory management to meet the diverse needs of its customer base.
Dataset overview
The fashion e-commerce store collects image data representing various fashion items, categorized into different classes, from its customers. The Fashion-MNIST dataset comprises 70,000 grayscale images of clothing and accessories, each associated with a specific label indicating its category and of size 28x28.
Features in the dataset include:
- Image data: Grayscale...
