Thank you for holding this competition.
What order is a column of “prev_items”?
a lower index is newer? or older?
Thank you for holding this competition.
What order is a column of “prev_items”?
a lower index is newer? or older?
Will it make any difference?
Are you planning to give more weightage to later ones or so?
I have the same question!
I want to truncate the session data. But don’t know the order
I have the same question!
Based on my local validation, i believe that lower index in the provided prev_items
list is older in time. And the larger index in prev_items
is newer in time. (Using this assumption results in better CV score).
Thanks! I totally agree with Chris.
I experiment time-weight method in Local Validation.
high index number is better than lower.
also I found this sentence in AIcrowd | Task 1 - Getting Started | Posts
The Multilingual Shopping Session Dataset is a collection of anonymized customer sessions containing products from six different locales, namely English, German, Japanese, French, Italian, and Spanish. It consists of two main components: user sessions and product attributes. User sessions are a list of products that a user has engaged with in chronological order, while product attributes include various details like product title, price in local currency, brand, color, and description.