Global optimization of inventory allocation

Abstract

A plurality of desirability prediction values are determined by one or more machine learning models. A desirability prediction value of the plurality of desirability prediction values corresponds to a particular client and a particular product. A plurality of global constraints are determined. A plurality of products are allocated to a plurality of clients based on the plurality of determined desirability prediction values and the plurality of determined global constraints.

What is claimed is: 
     
         1 . A method, comprising:
 determining, by one or more processors implementing one or more machine learning models, a plurality of desirability prediction values, wherein a desirability prediction value of the plurality of desirability prediction values corresponds to a particular client and a particular product;   determining, by the one or more processors, a plurality of global constraints; and   allocating, by the one or more processors, a plurality of products to a plurality of clients based on the plurality of determined desirability prediction values and the plurality of determined global constraints.   
     
     
         2 . The method of  claim 1 , wherein the desirability prediction value indicates a likelihood that the particular customer is to purchase the particular product. 
     
     
         3 . The method of  claim 1 , wherein the desirability prediction value is a match score for the particular client and the particular product. 
     
     
         4 . The method of  claim 1 , wherein the desirability prediction value is normalized. 
     
     
         5 . The method of  claim 1 , wherein the plurality of global constraints includes an inventory constraint. 
     
     
         6 . The method of  claim 5 , wherein the inventory constraint constrains the allocation of the plurality of products to a particular inventory metric. 
     
     
         7 . The method of  claim 5 , wherein the inventory constraint is based at least one of a current inventory, a certain time window, and/or a future inventory. 
     
     
         8 . The method of  claim 1 , wherein the plurality of constraints includes a constraint to limit a number of products made available to each of the plurality of clients. 
     
     
         9 . The method of  claim 1 , wherein the plurality of constraints includes a minimum viable assortment constraint. 
     
     
         10 . The method of  claim 1 , wherein the plurality of determined desirability prediction values are approximate desirability prediction values. 
     
     
         11 . The method of  claim 1 , wherein the plurality of determined global constraints are relaxed. 
     
     
         12 . The method of  claim 1 , wherein at least one of the plurality of global constraints is dropped. 
     
     
         13 . The method of  claim 1 , wherein at least one of the plurality of global constraints is approximated. 
     
     
         14 . The method of  claim 1 , further comprising providing to a reviewer a listing of the plurality of products allocated to the particular client. 
     
     
         15 . The method of  claim 14 , further comprising receiving from the reviewer an identification of a subset of the plurality of products allocated to the particular client. 
     
     
         16 . The method of  claim 15 , further comprising providing the identified subset of the plurality of products to the particular client. 
     
     
         17 . The method of  claim 16 , further comprising receiving feedback regarding at least one of the plurality of products and using the feedback to adjust the desirability prediction value for the particular product. 
     
     
         18 . A system, comprising:
 a memory; and   one or more processors coupled to the memory, wherein the one or more processors are configured to:
 determine, using one or more machine learning models, a plurality of desirability prediction values, wherein a desirability prediction value of the plurality of desirability prediction values corresponds to a particular client and a particular product; 
 determine a plurality of global constraints; and 
 allocate a plurality of products to a plurality of clients based on the plurality of determined desirability prediction values and the plurality of determined global constraints. 
   
     
     
         19 . The system of  claim 18 , wherein the desirability prediction value indicates a likelihood that the particular customer is to purchase the particular product. 
     
     
         20 . A computer program product, the computer program product being embodied in a non-transitory computer readable storage medium and comprising computer instructions for:
 determining, by implementing one or more machine learning models, a plurality of desirability prediction values, wherein a desirability prediction value of the plurality of desirability prediction values corresponds to a particular client and a particular product;   determining a plurality of global constraints; and   allocating a plurality of products to a plurality of clients based on the plurality of determined desirability prediction values and the plurality of determined global constraints.