# US 11,113,751 B2 · Systems and methods for predicting lost demand using machine learning architectures > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,113,751 B2 - Document ID: US11113751B2 - Status: Granted - Kind: B2 - Application number: 16/774,407 - Filed: 2020-01-28 - Published: 2021-07-29 - Granted: 2021-09-07 - Earliest claimed: 2020-01-28 - Assignees: Walmart Apollo, LLC - Applicants: Walmart Apollo, LLC - Inventors: Shengyang Zhang; Mingang Fu; Madhavan Kandhadai Vasantham - Web page: https://patentgrep.com/patent/US11113751B2 ## CPC - G06Q 30/0635: {replenishment orders; recurring orders} - G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound - G06N 20/00: Machine learning - G06N 20/20: Ensemble learning - G06Q 10/087: Inventory or stock management, e.g. order filling, procurement or balancing against orders - G06Q 10/0877: {by inventory control or reporting using inventory tracking or counting} - G06Q 30/0201: Market modelling; Market analysis; Collecting market data - G06Q 30/0202: Market predictions or forecasting for commercial activities - G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract Systems and methods including one or more processors and one or more non-transitory storage devices storing computing instructions configured to run on the one or more processors and perform acts of: providing, via an electronic platform, access to one or more order placement user interfaces; collecting order placement information associated with the one or more order placement user interfaces; analyzing, by a conversion determination network of a machine learning architecture, the order placement information; generating actual conversion information for client sessions based on the actual availability of the order placement options during the client sessions; generating predicted conversion information for the client sessions based on a full availability of all of the order placement options during the client sessions; and generating lost demand information based, at least in part, on the actual conversion information and the predicted conversion information. Other embodiments are disclosed herein. ## Claims ### Claim 1 A system comprising: one or more processors; and one or more non-transitory computer-readable storage devices storing computing instructions configured to run on the one or more processors and perform: providing, via an electronic platform, access to one or more order placement user interfaces; collecting order placement information associated with the one or more order placement user interfaces, the order placement information at least indicating (a) actual availability of order placement options during client sessions involving the one or more order placement user interfaces and (b) whether or not the client sessions were converted to actual orders; analyzing, by a conversion determination network of a machine learning architecture, the order placement information; generating, using the conversion determination network of the machine learning architecture, actual conversion information for the client sessions based on the actual availability of the order placement options during the client sessions; generating, using the conversion determination network of the machine learning architecture, predicted conversion information for the client sessions based on a full availability of all of the order placement options during the client sessions; and generating lost demand information based, at least in part, on the actual conversion information and the predicted conversion information. ### Claim 2 (depends on 1) The system of claim 1, wherein: the actual conversion information comprises a conversion probability for each of the client sessions based on the actual availability of the order placement options during the client sessions; and the predicted conversion information comprises a predicted conversion probability for each of the client sessions in hypothetical scenarios in which there is full availability of all of the order placement options during the client sessions. ### Claim 3 (depends on 2) The system of claim 2, wherein generating the lost demand information comprises: aggregating the conversion probabilities to produce a total conversion probability value; aggregating the predicted conversion probabilities to produce a total predicted conversion probability value; and generating the lost demand information by determining a difference between the total conversion probability value and the total predicted conversion probability value. ### Claim 4 (depends on 1) The system of claim 1, wherein the computing instructions are configured to run on the one or more processors and further perform: adjusting settings associated with the order placement options based on the lost demand information. ### Claim 5 (depends on 4) The system of claim 4, wherein adjusting the settings associated with the order placement options at least includes increasing availability of the order placement options in undersupply scenarios. ### Claim 6 (depends on 4) The system of claim 4, wherein adjusting the settings associated with the order placement options at least includes increasing labor resources. ### Claim 7 (depends on 1) The system of claim 1, wherein the computing instructions are configured to run on the one or more processors and further perform: pre-processing the order placement information to consolidate information relating to the client sessions. ### Claim 8 (depends on 7) The system of claim 7, wherein consolidating the information relating to the client sessions includes: identifying users that engaged in multiple client sessions involving the one or more order placement user interfaces within a pre-determined time period, wherein the client sessions comprise the multiple client sessions; selecting an initial client session from the multiple client sessions to be included in the order placement information; and deleting any other client sessions included in the multiple client sessions before analyzing the order placement information such that analyzing the order placement information excludes analyzing information relating to the other client sessions, and such that the actual conversion information excludes information relating to the other client sessions, and such that the predicted conversion information excludes information relating to the other client sessions. ### Claim 9 (depends on 1) The system of claim 1, wherein the order placement options correspond to at least one of: delivery options or pick-up options. ### Claim 10 (depends on 9) The system of claim 9, wherein generating the lost demand information includes at least of one: generating the lost demand information corresponding to the delivery options; generating the lost demand information corresponding to the pick-up options; or generating the lost demand information corresponding to both the delivery options and the pick-up options. ### Claim 11 A method implemented via execution of computing instructions configured to run at one or more processors and configured to be stored at non-transitory computer-readable media, the method comprising: providing, via an electronic platform, access to one or more order placement user interfaces; collecting order placement information associated with the one or more order placement user interfaces, the order placement information at least indicating (a) actual availability of order placement options during client sessions involving the one or more order placement user interfaces and (b) whether or not the client sessions were converted to actual orders; analyzing, by a conversion determination network of a machine learning architecture, the order placement information; generating, using the conversion determination network of the machine learning architecture, actual conversion information for the client sessions based on the actual availability of the order placement options during the client sessions; generating, using the conversion determination network of the machine learning architecture, predicted conversion information for the client sessions based on a full availability of all of the order placement options during the client sessions; and generating lost demand information based, at least in part, on the actual conversion information and the predicted conversion information. ### Claim 12 (depends on 11) The method of claim 11, wherein: the actual conversion information comprises a conversion probability for each of the client sessions based on the actual availability of the order placement options during the client sessions; and the predicted conversion information comprises a predicted conversion probability for each of the client sessions in hypothetical scenarios in which there is full availability of all of the order placement options during the client sessions. ### Claim 13 (depends on 12) The method of claim 12, wherein generating the lost demand information comprises: aggregating the conversion probabilities to produce a total conversion probability value; aggregating the predicted conversion probabilities to produce a total predicted conversion probability value; and generating the lost demand information by determining a difference between the total conversion probability value and the total predicted conversion probability value. ### Claim 14 (depends on 11) The method of claim 11 further comprising: adjusting settings associated with the order placement options based on the lost demand information. ### Claim 15 (depends on 14) The method of claim 14, wherein adjusting the settings associated with the order placement options at least includes increasing availability of the order placement options in undersupply scenarios. ### Claim 16 (depends on 14) The method of claim 14 further comprising: adjusting settings associated with labor resources based on the lost demand information. ### Claim 17 (depends on 11) The method of claim 11 further comprising: pre-processing the order placement information to consolidate information relating to the client sessions. ### Claim 18 (depends on 17) The method of claim 17, wherein consolidating the information relating to the client sessions includes: identifying users that engaged in multiple client sessions involving the one or more order placement user interfaces within a pre-determined time period, wherein the client sessions comprise the multiple client sessions; selecting an initial client session from the multiple client sessions to be included in the order placement information; and deleting any other client sessions included in the multiple client sessions before analyzing the order placement information such that analyzing the order placement information excludes analyzing information relating to the other client sessions, and such that the actual conversion information excludes information relating to the other client sessions, and such that the predicted conversion information excludes information relating to the other client sessions. ### Claim 19 (depends on 11) The method of claim 11, wherein the order placement options correspond to at least one of: delivery options or pick-up options. ### Claim 20 (depends on 19) The method of claim 19, wherein generating the lost demand information includes at least of one: generating the lost demand information corresponding to the delivery options; generating the lost demand information corresponding to the pick-up options; or generating the lost demand information corresponding to both the delivery options and the pick-up options.