From the course: Data Planning, Strategy, and Compliance for AI Initiatives
Join today to access over 25,500 courses taught by industry experts.
Solution: Engineer features
From the course: Data Planning, Strategy, and Compliance for AI Initiatives
Solution: Engineer features
(uplifting music) - [Instructor] Now, there are many possible examples of engineered features in a retail dataset, but here are some. We could look at average days between purchases, or the purchase frequency. We could look at days since last purchase or the recency of the latest purchase. Also, we can look at average basket size, so how many items are people actually buying when they go to a store? We could also look at basket diversities, so how many different categories are represented in a basket or in a transaction? We could also look at some temporal features like day of week or time of day purchasing patterns as well. And we could consider category affinity scores or preference for certain departments over others. These are all examples of engineered features related to retail data sets.
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.