The data provided contains the number of items sold of each item (or SKU) each month. The dataset also provides some details as to the number of items sold through each major channel. Channels are Fossil’s customers. A channel can be an online retailer, an in-person retail chain store, etc. There are other minor channels that are not represented in the dataset as a channel, but the total sales represent all sales across the major channels as well as smaller channels.
The data provided also contains inventory data such as the beginning of period (BOP), or starting, inventory (inventory present at the beginning of the month) and on-hand, or customer side inventory (inventory available through customers or channels).
The objective is to accurately predict the “sellin”, which is the optimal inventory needed to meet demand by Fossil to its own customers or channels, four months into the future.
This file resembles Train.csv but without the target-related columns. This is the dataset on which you will apply your model to.
This file contains the target. This is the dataset that you will use to train your model.
This file explains the variables.
Shows the submission format for this competition, with the ‘Item ID’ column containing all item IDs for the test set and columns containing your probabilistic predictions for the different crop types. The order of the rows does not matter, but name of the item ID needs to be correct.