Quantity Prediction Algorithm

Asked Sep 07 '15 at 15:16

Active Sep 07 '15 at 15:16

Viewed 464 times

I want to make prediction for quantity of stock that will be sufficient over a period of time i.e from one delivery to another.

Assuming, i want a store to have minimum amount of item between deliveries e.g 30 items, delivery interval can be 7 , 14, 30, days.

Also i have history data of past delivery interval, amount of item delivered and the amount of item available before next delivery.

What will be the best way to relate these variable: delivery interval and on hand quantity to quantity delivered.

also can i predict quantity to be delivered in other to get a given amount of on hand stock before next delivery.

asked Sep 07 '15 at 15:16

This is a function fitting problem. Do you have data to share ? This might suggest which form of functional approximation might be most suitable. If not, why not start with a linear regression and move on from there to add terms in $x^n$. – image_doctor Sep 08 '15 at 07:51
@image_doctor thanks, actually that's what i decided to use, Linear Regression, while i observe how that work out. – Jideobi Benedine Ofomah Sep 08 '15 at 09:06
Excellent, let us know how it goes, visual inspection of the fit you get can give a good indication of how well your regression model is working. – image_doctor Sep 08 '15 at 10:04
Consider Poisson regression since you have count data. – Emre Sep 08 '15 at 19:47

0 Answers0