Abstract: In some examples, a system may be configured to obtain a set of features of a set of users including one or more features of transaction of the set of users and one or more features of engagement data of the set of users. Additionally, the system may be configured to implement a first set of operations that generate output data including a plurality of churn scores, based on the set of features. In some examples, each churn score of the plurality of churn scores being associated with a particular user of the set of users and characterize a likelihood of a churn event of the corresponding user.

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for electric grid modeling and simulation. One of the methods includes, obtaining a model of an electrical power system; executing a simulation of the model including by separately simulating, in parallel, behaviors of the electric power system during each of a consecutive series of time periods to provide individual simulation results for each time period, where the consecutive series of time periods together represent a simulation time frame over which the behaviors of the electric power system are simulated; combining the individual simulation results into a simulation output for the simulation time frame; and providing, for display, the simulation output.

Abstract: A system is configured to train a customer understanding model to generate a preference score for substitution items. The customer understanding model generates a preference score for each of a plurality of related substitution items based on order data including data indicative of at least one item ordered and location data indicating a location of a first store. The customer understanding model ranks each of the substitution items based on the preference score. Order data is transmitted including substitution data identifying each of the substitution items and corresponding rank. Performance data associated with a set of operations implemented based on the order data and the substitution data is obtained. An updated customer understanding model is trained based on the performance data and iteratively modified based on the updated training dataset and updated performance metrics generated from second performance data.

Abstract: Systems and methods for generating and using seasonal affinity scores is disclosed. A set of user-specific historical transaction data is obtained and a user-specific affinity score including at least one of a user-specific season affinity score or a user-specific seasonal theme affinity score is determined by determining one or more product affinity scores for a set of product taxonomies and combining the one or more product affinity scores with one or more product index scores to generate the user-specific affinity score. The product affinity scores are determined by a trained scoring calculation model configured to receive the set of user-specific historical transaction data. One or more interface elements are selected based on the user-specific affinity score and an interface is generated including the one or more interface elements.

Abstract: Systems and methods for context aware engagement are disclosed. A request for a user interface, including a user identifier, is received. A set of features associated with the user identifier are obtained and a user embedding is generated by applying an autoencoder to the set of features. A set of potential tasks associated with an enrollment portion of the user interface is obtained. A task embedding is generated for each task in the set of potential tasks. A user-task affinity is generated by comparing the user embedding to each task embedding. A ranked set of tasks is generated by ranking each task based on the user-task affinity. A set of interface elements related to the highest ranked tasks in the ranked set of tasks is generated. A user interface including interface elements is generated and transmitted to a device that requested the user interface.

Abstract: Systems and methods for hybrid optimization of training ranking models is disclosed. A training dataset including a plurality of anchor items, a plurality of recommended item sets, and ground truth data is obtained from a database. A base machine learning model including a step function configured to determine a relevance score is iteratively trained to generate a trained ranking model. The plurality of anchor items and the plurality of recommended item sets are provided as an input to the base machine learning model and the ground truth is provided as a target output. The step function is trained using an adaptive step size according to a first order Barzilai-Borwein (BB) method and a line search method. The trained ranking model is stored in non-transitory memory.

Abstract: A system including one or more processors and one or more non-transitory computer-readable media storing computing instructions that, when executed on the one or more processors, cause the one or more processors to perform operations: generating, using a trained machine-learning model, personalized product-type metrics for a user based on historic activity of the user and product-type pairs in an item taxonomy; determining top product types for the user based on an anchor item; determining a set of first items associated with the top product types; ranking each item in the set of first items for (i) the anchor item and (ii) for each item in the set of first items; and selecting, based on the ranking, a set of top items from the set of first items to be personalized complementary item recommendations for the user based on the anchor item. Other embodiments are described.

Abstract: Methods, systems, and apparatus, including medium-encoded computer program products, for efficient storage for electrical grid models, which can include the actions of obtaining a graph storage structure that can include component nodes, each component node representing a component of an electric grid within a computer simulatable model and which can include an identity of the component, wherein each component node is connected to one or more edges, and wherein each edge can include a version identifier, each edge connected to: (i) a second component node or (ii) a value node; obtaining graph update data that can include changes to at least one component; updating the graph storage structure according to the graph update data; and storing at least a subset of the graph update data in association with at least one new edge that includes information indicating the update data is relevant to a new version of the model.

Abstract: Systems and methods for recommending items based on enhanced user representations are disclosed. A sparse part and a dense part of user-item interaction data are generated. While the dense part is split into a plurality of training data batches, the sparse part is split into a plurality of inference data batches. A deep learning model is trained based on the plurality of training data batches. Inferred user embeddings are generated by applying the trained deep learning model to the plurality of inference data batches in parallel. The inferred user embeddings are non-zero user representations in a same latent space. Based on user session data of a query user and the inferred user embeddings, recommended items are generated and transmitted to a user device for display to the query user.

Abstract: Disclosed herein is a method of provisioning a virtual experience based on user preference. The method may include receiving an identity data associated with an identity of a user, retrieving a user profile data based on the identity data, analyzing the user profile data using a machine learning model, determining at least one preference data based on the analyzing, generating an interactive 3D model data, transmitting the interactive 3D model data to a user device configured to present the interactive 3D model data, receiving a reaction data from the user device. Further, the user device may include at least one sensor configured to generate the reaction data based on a behavioral reaction of a user consuming the interactive 3D model data.

Abstract: In some examples, at least one processor executes instructions to, during a chat session, obtain, from a computing device of a first user, session data associated with the chat session and a first user, obtain a first set of data identifying a set of price drop items associated with the chat session, and obtain a second set of data identifying a set of attribute values associated with one or more attribute features of each item type. Based on the first set of data and the second set of data, a third set of data including affinity value data for each price drop item is generated. A fourth set of data identifying a set of items each with a score is obtained. A fifth set of data is then generated based on the third set of data and the fourth set of data.

Abstract: Systems and methods for post transaction seasonal item recommendations are disclosed. In some embodiments, a current seasonal time window associated with a seasonal event and some seasonal product types is determined. Based on historical transaction data of the seasonal product types, a first seasonal index score is computed for each item, and a second seasonal index score is computed for each product type including one or more items. A seasonal rank score is generated for each item based on the first seasonal index score and the second seasonal index score, such that the items in the historical transaction data are ranked based on their respective seasonal rank scores. Based on the ranked items and a transaction order from a user, a list of recommended items is generated and displayed to the user.

Abstract: Systems and methods of generating an interface including cross-pollinated interface elements are disclosed. A request for an interface for a first intent is received. The request includes a user identifier. An interface generation engine generates an interface including first items associated with the first intent and cross-pollinated items associated with a second intent. The set of cross-pollinated items are selected based on a cross-pollination score. The interface generation engine inserts the items into the interface and transmits the interface to a user device associated with the user identifier. A cross-pollination engine generates the cross-pollination score using a trained sequential prediction model configured to receive the set of features associated with the user identifier and output the cross-pollination score. The cross-pollination score represents a likelihood of a user associated with the user identifier interacting with at least one cross-pollinated item.

Abstract: Systems and methods of generating an interface including elements related to a next best state prediction are disclosed. A request for an interface including a user identifier is received. A next state prediction engine receives a sequence unit set including at least one sequence unit associated with the user identifier and a set of features associated with the at least one sequence unit and generates at least one next state prediction using a trained sequential prediction model. The trained sequential prediction model is configured to receive the sequence unit set and the set of features for the at least one sequence unit and output at least one predicted next state for the sequence unit set. An interface generation engine generates an interface including at least one element related to the at least one predicted next state and transmits the interface to a user device associated with the user identifier.

Abstract: Systems and methods of generating an interface including one or more assets selected by an asset prediction model are disclosed. A user identifier associated with a set of user features and a set of assets each including a set of asset features is received and a set of predicted assets is generated using a trained asset prediction model. The trained asset prediction model comprises a machine learning model configured to receive the set of user features and the set of asset features for each asset in the set of assets and output the set of predicted assets and the trained asset prediction model is configured to maximize a likelihood of engagement for the set of predicted asset. An interface including a predetermined number of assets selected from the set of predicted assets in descending ranked order is generated.

Abstract: A privacy system includes a computing device configured to obtain user transactional data characterizing at least one transaction of a user on an ecommerce marketplace and to determine a privacy vulnerability score of the user by comparing the transactional data to a user vulnerability distribution. The computing device is also configured to send the privacy vulnerability score to a personalization engine.

Abstract: Systems and methods including one or more processing modules and one or more non-transitory storage modules storing computing instructions configured to run on the one or more processing modules and perform acts of accessing first transaction data stored in a transaction database, the first transaction data describing first transactions for first items from first users; determining, using the first transaction data, first micro-intents associated with the first transaction data; grouping the first micro-intents into clusters; labeling each cluster of the first micro-intents with a respective label; receiving second transaction data of a user, the second transaction data describing second transactions for second items for the user; determining, using the second transaction data, second micro-intents present in the second transactions; receiving current transaction data from a user interface of an electronic device of the user; determining, using the current transaction data, that the user is expressing a curre

Abstract: A method including training a recurrent neural network model to create a trained model based at least on: (a) first images associated with first items on a website, (b) first search terms used by users of the website to search for the first items on the website, and (c) personal features of the users. The method also can include receiving an input image that was uploaded by a current user, the input image comprising a depiction of one or more items. The method additionally can include obtaining a user encoded representation vector for the current user based on a set of personal features of the current user. The method further can include generating an image encoded representation vector for the input image.

Abstract: Systems and methods utilizing a classification model and a ranking model are disclosed. A user identifier is received and a user profile is generated. The user profile includes a plurality of user features. The user profile is classified into a classification using a trained classification model. The trained classification model receives a first subset of the plurality of user features. A set of communication elements is ranked using a trained ranking model. The trained ranking model receives a second subset of the plurality of user features. An electronic communication including a plurality of interface elements is generated. The plurality of interface elements includes at least one communication element selected from the ranked set of communication elements in descending ranked order. A type of the electronic communication is selected based on the classification of the user profile.

Abstract: Systems and methods for providing noise-resistant complementary item recommendations are disclosed. A trained model is generated based on transaction data to represent each item of a set of items as a Gaussian distribution with a mean vector and a non-zero covariance matrix. An anchor item is to be displayed to a user via a user interface executed on a user device of the user, and is represented as a Gaussian distribution with an anchor mean vector and an anchor non-zero covariance matrix. A complementarity score for each item is computed based on a distance between the mean vector of the item and the anchor mean vector to generate a ranking for the set of items based on their respective complementarity scores. A plurality of top items are selected from the set of items based on the ranking as recommended complementary items, which are displayed with the anchor item on the user interface.