System and method for retail price optimization

A method of generating an electronic retail price model and a selected control policy, the electronic retail price model and the selected control policy for generating an electronic retail price plan for retail operations performed within an enterprise network comprising a plurality of distributed computing devices, the plurality of distributed computing devices comprising a plurality of point-of-sale devices and a plurality of enterprise servers, the plurality of distributed computing devices in network communication, comprising:

• receiving, at a first server of the plurality of enterprise servers, transaction data from the plurality of point-of-sale devices, the transaction data defined as a time series including a plurality of transaction data items corresponding to one or more previous time periods, the plurality of transaction data items comprising at least one of price data, product data, and customer data;
• providing, at the first server, an intelligent agent configured to operate in an operational control loop and a simulated control loop, the intelligent agent comprising:
• a memory component comprising a short-term memory, the short-term memory storing one or more simulated system states and one or more simulated system rewards, and a long-term memory, the long-term memory storing a plurality of long-term sensor data;
• and
• a simulation component to simulate an operational system of the operational control loop;
• determining, by the intelligent agent, transformed transaction data including a power spectrum of the time series transaction data, the power spectrum comprising a vertical axis of amplitude and a horizontal axis of frequency and the transformed transaction data indicating gross or aberrant trends in the transaction data;
• determining, by the intelligent agent, one or more features from the transformed transaction data;
• determining, by the intelligent agent, one or more actionable features from the one or more features;
• generating an electronic retail price model of the intelligent agent from the one or more actionable features;
• selecting a selected control policy for the electronic retail price model;
• executing a reinforcement learning algorithm to simulate, at the simulation component of the intelligent agent, a plurality of simulated pricing decisions based on the electronic retail price model and the selected control policy to update the one or more simulated system states and the one or more simulated system rewards;
• receiving, at the intelligent agent, the updated one or more simulated system states and the updated one or more simulated system rewards from the simulation component of the intelligent agent;
• receiving, at the intelligent agent, a measured state of the operational system and a measured reward of the operational system;
• correcting, using the reinforcement learning algorithm, the one or more simulated system states based on the measured state of the operational system and the measured reward of the operational system;
• and
• determining, by the intelligent agent operating in the operational control loop, the electronic retail price plan based on the one or more simulated system states and the one or more simulated system rewards.