# US 2020/0401880 A1 · Generating a recommended target audience based on determining a predicted attendance utilizing a machine learning approach > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2020/0401880 A1 - Document ID: US20200401880A1 - Status: Published application - Kind: A1 - Application number: 16/446,386 - Filed: 2019-06-19 - Published: 2020-12-24 - Earliest claimed: 2019-06-19 - Assignees: Adobe Inc. - Applicants: Adobe Inc. - Inventors: Niranjan Kumbi; Vaidyanathan Venkatraman; Rajan Madhavan; Omar Rahman; Kai Lau; Badsah Mukherji; Ajay Awatramani - Web page: https://patentgrep.com/patent/US20200401880A1 ## CPC - G06N 3/08: Learning methods - G06N 3/045: Combinations of networks - G06N 3/0464: Convolutional networks [CNN, ConvNet] - G06N 3/084: Backpropagation, e.g. using gradient descent - G06N 3/09: Supervised learning - G06N 3/0985: Hyperparameter optimisation; Meta-learning; Learning-to-learn - G06Q 10/04: Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" (market predictions or forecasting for commercial activities G06Q30/0202) - G06Q 30/0202: Market predictions or forecasting for commercial activities - G06N 3/048: Activation functions - G06N 20/00: Machine learning - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - 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 ## Abstract The present disclosure relates to systems, methods, and non-transitory computer readable media for generating a recommended target audience based on determining a predicted attendance utilizing a neural network approach. For example, the disclosed systems can utilize an approximate nearest neighbor algorithm to identify individuals that are within a similarity threshold of invitees for an event. In addition, the disclosed systems can implement an attendance prediction model to determine a probability of an invitee attending the event. The disclosed systems can further determine a predicted attendance for an event based on the individual probabilities. Based on identifying the similar individuals to, and the attendance probabilities for, the invitees, the disclosed systems can generate a recommended target audience to satisfy a target attendance for an event based on a predicted attendance for the event. ## Claims ### Claim 1 In a digital medium environment for organizing participation in events by distributing digital communications across computer networks, a computer-implemented method of generating a target audience, the computer-implemented method comprising: utilizing a nearest neighbor algorithm to identify, based on features associated with a plurality of individuals, one or more individuals that satisfy a similarity threshold with respect to a set of invitees for an event; a step for determining a predicted attendance for the event by utilizing an attendance prediction model; and generating a recommended target audience to achieve a target attendance for the event based on the predicted attendance for the event. ### Claim 2 (depends on 1) The computer-implemented method of claim 1, further comprising: receiving an indication to prioritize a feature in utilizing the nearest neighbor algorithm to identify the one or more individuals that satisfy the similarity threshold; and weighting the feature more heavily within the nearest neighbor algorithm. ### Claim 3 (depends on 2) The computer-implemented method of claim 2, wherein utilizing the nearest neighbor algorithm comprises utilizing an approximate nearest neighbor algorithm. ### Claim 4 (depends on 1) The computer-implemented method of claim 1, further comprising utilizing hyperparameter tuning to determine a number of nearest neighbor values for the nearest neighbor algorithm based on analyzing the plurality of individuals. ### Claim 5 (depends on 1) The computer-implemented method of claim 1, further comprising distributing electronic communications to client devices associated with the recommended target audience to attend the event. ### Claim 6 (depends on 1) The computer-implemented method of claim 1, wherein the attendance prediction model comprises a deep neural network. ### Claim 7 A non-transitory computer readable medium comprising instructions that, when executed by at least one processor, cause a computer device to: utilize a nearest neighbor algorithm to identify, based on features associated with a plurality of individuals, one or more individuals that satisfy a similarity threshold with respect to a set of invitees for an event; generate, by applying an attendance prediction model to the set of invitees for the event, probabilities that the invitees within the set will attend the event; determine a predicted attendance for the event based on the probabilities that the invitees within the set will attend the event; and generate, based on the one or more individuals that satisfy the similarity threshold, a recommended target audience for the event to achieve, based on the predicted attendance, a target attendance for the event. ### Claim 8 (depends on 7) The non-transitory computer readable medium of claim 7, further comprising instructions that, when executed by the at least one processor, cause the computer device to: receive an indication to prioritize a feature in utilizing the nearest neighbor algorithm to identify the one or more individuals that satisfy the similarity threshold; and in response to the indication to prioritize the feature, weight the feature more heavily within the nearest neighbor algorithm. ### Claim 9 (depends on 7) The non-transitory computer readable medium of claim 7, further comprising instructions that, when executed by the at least one processor, cause the computer device to: generate, by applying the attendance prediction model to the one or more individuals that satisfy the similarity threshold, attendance probabilities that the one or more individuals will attend the event; and rank the one or more individuals that satisfy the similarity threshold based on the attendance probabilities. ### Claim 10 (depends on 9) The non-transitory computer readable medium of claim 9, further comprising instructions that, when executed by the at least one processor, cause the computer device to generate the recommended target audience by selecting an individual from the one or more individuals that satisfy the similarity threshold based on the ranking. ### Claim 11 (depends on 9) The non-transitory computer readable medium of claim 9, further comprising instructions that, when executed by the at least one processor, cause the computer device to determine the predicted attendance for the event by combining the probabilities that the invitees within the set will attend the event. ### Claim 12 (depends on 7) The non-transitory computer readable medium of claim 7, further comprising instructions that, when executed by the at least one processor, cause the computer device to train the attendance prediction model to generate the probabilities that the invitees will attend the event based on training features and ground truth attendance information. ### Claim 13 (depends on 12) The non-transitory computer readable medium of claim 12, wherein the attendance prediction model comprises a deep neural network. ### Claim 14 A system comprising: at least one processor; and a non-transitory computer readable medium comprising instructions that, when executed by the at least one processor, cause the system to: utilize an approximate nearest neighbor algorithm to identify, based on features associated with a plurality of individuals, one or more similar individuals that satisfy a similarity threshold with respect to a set of invitees for an event; generate probabilities that the invitees within the set will attend the event; generate attendance probabilities that the one or more similar individuals will attend the event; determine a predicted attendance for the event based on the probabilities that the invitees within the set will attend the event by summing the probabilities that the invitees within the set will attend the event; rank the one or more similar individuals that satisfy the similarity threshold based on the attendance probabilities that the one or more similar individuals will attend the event; and generate, based on the ranking of the one or more similar individuals, a recommended target audience for the event to achieve, based on the predicted attendance, a target attendance for the event. ### Claim 15 (depends on 14) The system of claim 14, further comprising instructions that, when executed by the at least one processor, cause the system to determine a difference between the predicted attendance for the event and the target attendance. ### Claim 16 (depends on 15) The system of claim 15, further comprising instructions that, when executed by the at least one processor, cause the system to generate the recommended target audience by selecting, from the ranking, a number of similar individuals with associated attendance probabilities that sum to at least the difference between the predicted attendance for the event and the target attendance. ### Claim 17 (depends on 14) The system of claim 14, further comprising instructions that, when executed by the at least one processor, cause the system to: receive an indication to prioritize a feature of the plurality of features in utilizing the approximate nearest neighbor algorithm to identify the one or more individuals that satisfy the similarity threshold; and in response to the indication to prioritize the feature, weight the feature more heavily within the approximate nearest neighbor algorithm. ### Claim 18 (depends on 14) The system of claim 14, further comprising instructions that, when executed by the at least one processor, cause the system to: utilize the approximate nearest neighbor algorithm to identify the one or more similar individuals that satisfy the similarity threshold based on a first set of features; and generate the probabilities that the invitees within the set will attend the event by applying the attendance prediction model based on a second set of features. ### Claim 19 (depends on 14) The system of claim 14, further comprising instructions that, when executed by the at least one processor, cause the system to distribute electronic communications to client devices associated with the recommended target audience to attend the event. ### Claim 20 (depends on 14) The system of claim 14, further comprising instructions that, when executed by the at least one processor, cause the system to train the attendance prediction model to generate the probabilities that the invitees within the set will attend the event based on training features and ground truth attendance information, wherein the attendance prediction model comprises a deep neural network.