Would really appreciate if anyone happen to know h...
# general
t
Would really appreciate if anyone happen to know how handle missing dates in ts
m
The classical way for sparse date is to create those missing dates and fill them with 0s given that on those date nothing happened.
Instead of: Date 1: 234 Date 5: 7787 you would have: Date 1: 234 Date 2: 0 Date 3: 0 Date 4: 0 Date 5: 7787
t
And apply to model won’t it affect the model performance?
m
Yes, it will. But for the better, assuming that missing data indeed signifies 0. This is often the case in retail data: stores don't usually report 0 sales for items or SKUs. If something doesn't sell, they simply omit that information.