1
$\begingroup$

Is it possible to predict the products which are going to be out of stock in next coming days/weeks.?

If Yes, what could be the possible data points that I should collect?

How is this problem different from sales prediction?

| improve this question | |
$\endgroup$
  • 1
    $\begingroup$ Hi can you be more specific on what input data you have? $\endgroup$ – El Burro Apr 21 '18 at 17:43
  • $\begingroup$ @ElBurro I have product data like product id, price, status(1=in stock,0=out of stock),product created date on website and last update about the product. can we go for time series to forecast the status ? or i have to go ML approach..? another data point that i can get is number of items in stock for each product.. $\endgroup$ – john Mohd Apr 23 '18 at 5:48
  • $\begingroup$ I *think that with the exception of simple statistics you will need more variables like price, size of stock, recent change history, product category etc. $\endgroup$ – El Burro Apr 23 '18 at 7:08
  • $\begingroup$ @ElBurro i have price, number of items available of a particular product and last updated time (when the item was put in stock last time). $\endgroup$ – john Mohd Apr 23 '18 at 11:39
1
$\begingroup$

You may don’t need any machine learning for that, just simple math.

Calculate the average use per day, define your desired minimum amount which should be on stock and collect the time for a supplier to send you new products.

After that, you can predict when to order your product from the supplier to always have your desired amount on stock.

Search for order point system :)

With the system you could estimate when a product will be out of stock and when to order to prevent that.

| improve this answer | |
$\endgroup$
  • $\begingroup$ thanks @MBDev... Its not seller specific. For example Alibaba.com has multiple sellers for each product, I want to predict when that product could be out of stock from all the sellers...which means product out of stock at Alibaba.com.. $\endgroup$ – john Mohd Apr 21 '18 at 10:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.