You will see a lot of #posts on #Datascience #alogrithms but very few posts on #features #data. 80% of core learning is #features and #data. Hope #features are also #shared more openly for benefit of the community and #learning from #experiements.
Pay deep attention to features. For example, For an average shopper what all I can track about his purchase?
- How many average views to buy a product
- How many competitive products comparison
- How many average price comparisons
- How many discount comparisons
- Time of purchase (Weekday / Weekend)
- Festive Shopper
- Bought together supplementary items
- Weekend shopper
- Late night shopper
- Salary Day Shopper
- Discount Shopper
Many times it is a tradeoff between 10000 feet view vs 1000 feet ground view. Balance both #perspective vs #bigpicture. Nth level personalization / real-time personalization across browsers/devices like we see how google recommends in the next browser session. #perspectives #datascience #persppectives #learningmoments
No comments:
Post a Comment