Price optimization with rule relaxation
View Patent ↗A price optimization system computes a preferred set of prices for a plurality of products. The optimization system stores a plurality of rules. The system may also prioritize the plurality of rules, and further relax lower priority rules to allow higher priority rules to become feasible. The feasible set of rules may then be utilized to generate the preferred set of prices.
1. Computer readable code, embodied on a computer readable media, for computing a preferred set of prices for a plurality of products, comprising an optimization engine comprising computer readable media, comprising:
computer readable code for storing a plurality of rules, wherein each rule includes constraints;
computer readable code for prioritizing the plurality of rules;
computer readable code for identifying an infeasible rule of the plurality of rules, wherein the infeasible rule is not able to be satisfied;
computer readable code for identifying at least one lower priority rule which has a lower priority than the infeasible rule, wherein the at least one lower priority rule conflicts with the infeasible rule;
computer readable code for incrementally relaxing the at least one lower priority rule to enable the infeasible rule to become feasible, wherein the incremental relaxing of the at least one lower priority rule includes broadening the constraints of the said at least one lower priority rule such that the conflict between the at least one lower priority rule and the infeasible rule is eliminated; and
computer readable code for performing an optimization process, utilizing Bayesian shrinkage modeling, to generate the preferred set of prices wherein the preferred set of prices meets the plurality of rules wherein a rule is feasible if said optimization process can be performed without violating said rule, further wherein the preferred set of prices is a set of product prices generated for said plurality of products.
2. The computer readable code according to claim 1 , wherein the computer readable code for relaxing at least one lower priority rule, comprises:
computer readable code for determining a priority of a rule determined to be infeasible;
computer readable code for determining a lowest priority infeasible rule;
computer readable code for determining if at least one rule with a lower priority than the priority of the rule determined to be the lowest priority infeasible rule may be relaxed to allow the rule determined to be the lowest priority infeasible rule to become feasible; and
computer readable code for incrementally relaxing at least one rule with a lower priority than the priority of the rule determined to be the lowest priority infeasible rule to enable the rule determined to be the lowest priority infeasible rule to become feasible, wherein incrementally relaxing said at least one rule comprises:
prioritizing each said at least one rule with a lower priority than the priority of the rule determined to be the lowest priority infeasible rule; and
sequentially relaxing in order of priority from lowest priority to highest priority each said at least one rule with a lower priority than the priority of the rule determined to be the lowest priority infeasible rule to a point that enables said lowest priority infeasible rule to become feasible, wherein each said rule is relaxed incrementally in sequence until said point is reached.
3. The computer readable code according to claim 1 , further comprising computer readable code for enabling a user to-edit the plurality of rules and prioritize the plurality of rules.
4. The computer readable code according to claim 1 , wherein at least one of the plurality of rules is a gross margin rule, which defines a constraint on gross margin changes.
5. The computer readable code according to claim 4 , wherein the constraint on change of the gross margin is placed on each product of a group of products.
6. The computer readable code according to claim 4 , wherein the constraint on change of the gross margin is placed on an average gross margin of a group of products.
7. The computer readable code according to claim 1 , wherein at least one of the plurality of rules is a store level volume rule, which defines a constraint on volume of sales changes at a store level.
8. The computer readable code according to claim 1 , wherein at least one of the plurality of rules is a competition rule, which provides a constraint on difference between at least one competitor's prices and the preferred set of prices.
9. A method for computing a preferred set of prices for a plurality of products, comprising generating a preferred set of prices in a computer system, comprising, the computer-implemented steps of:
storing a plurality of rules, wherein each rule includes constraints;
prioritizing the plurality of rules;
identifying an infeasible rule of the plurality of rules, wherein the infeasible rule is not able to be satisfied;
identifying at least one lower priority rule which has a lower priority than the infeasible rule, wherein the at least one lower priority rule conflicts with the infeasible rule;
incrementally relaxing the at least one lower priority rule to enable the infeasible rules to become feasible, wherein the incremental relaxing of the at least one lower priority rule includes broadening the constraints of the said at least one lower priority rule such that the conflict between the at least one lower priority rule and the infeasible rule is eliminated;
performing an optimization process, utilizing Bayesian shrinkage modeling, to generate the preferred set of prices wherein the preferred set of prices meets the plurality of rules wherein a rule is feasible if said optimization process can be performed without violating said rule, further wherein the preferred set of prices is a set of product prices generated for said plurality of products; and
setting prices for the plurality of products to the preferred set of prices, wherein the setting prices for the plurality of products includes displaying the preferred set of prices on a display on a computer system.
10. The method, as recited in claim 9 , wherein the relaxing at least one lower priority rule, comprises:
determining a priority of a rule determined to be infeasible;
determining a lowest priority infeasible rule;
determining if at least one rule with a lower priority than the priority of the rule determined to be the lowest priority infeasible rule may be relaxed to enable the rule determined to be the lowest priority infeasible rule to become feasible; and
incrementally relaxing at least one rule with a lower priority than the priority of the rule determined to be the lowest priority infeasible rule to enable the rule determined to be the lowest priority infeasible rule to become feasible, wherein incrementally relaxing said at least one rule comprises:
prioritizing each said at least one rule with a lower priority than the priority of the rule determined to be the lowest priority infeasible rule; and
sequentially relaxing in order of priority from lowest priority to highest priority each said at least one rule with a lower priority than the priority of the rule determined to be the lowest priority infeasible rule to a point that enables said lowest priority infeasible rule to become feasible, wherein each said rule is relaxed incrementally in sequence until said point is reached.
11. The method, as recited in claim 9 , further comprising enabling a user to edit the plurality of rules and prioritize the plurality of rules.
12. The method, as recited in claim 9 , wherein at least one of the plurality of rules is a gross margin rule, which defines a constraint on gross margin changes.
13. The method, as recited in claim 12 , wherein the constraint on change of the gross margin is placed on each product of a group of products.
14. The method, as recited in claim 12 , wherein the constraint on change of the gross margin is placed on an average gross margin of a group of products.
15. The method, as recited in claim 9 , wherein at least one of the plurality of rules is a store level volume rule, which defines a constraint on volume of sales changes at a store level.
16. The method, as recited in claim 9 , wherein at least one of the plurality of rules is a competition rule, which provides a constraint on difference between at least one competitor's prices and the preferred set of prices.