Fossil Demand Forecasting Challenge
Hi Zindi,
In the dataset, we have some time gaps of no sale records for already existing SKU's. Some SKU's are obsolete after sometime and 2020 is explained to be removed on purpose but we still have some mild time gaps for SKU's that continue till 2022. i.e From the beginning of first SKU record in database till an observable record in 2021, we have some gaps in the records within the timelines.
Since, the major objective here is to build a useful solution for Fossil, I will like to know if they need us to take this into account for our system to predict no sales for some timelines, rather than just output a predicted sale numbers everytime and also what these time gaps really mean.
Technically, I will think it's too much of an ideal situation that all SKU will always have sales, but it's possible.
Thank you for the clarification.
Hi @flamethrower, the only time gaps removed on purpose or missing are only for the year 2020; the rest means either the items went out of stock or the company stopped producing the items. In some cases, the item was out of stock like in June and July only to resurface in August. This means you won't have details for June and July in the dataset provided.
Best,
Zindi Team.
Hi @rwambu,
Thanks for this response. Much clearer.
Appreciate it.