Adobe, Inc.

Nikhil Sheoran

Atanu Ranjan Sinha

Deepali Jain

Deepali Gupta

Subrata Mitra

Sopan Khosla

Eunyee Koh

Sungchul Kim

Branislav Kveton

Hung Hai Bui

Dhruv Singal

Charles Li Chen

Nayan Raju Vysyaraju

Varun Srivastava

Nisheeth Golakiya

Shivakumar Vaithyanathan

Anup Bandigadi Rao

Tung Mai

Tong Yu

Sapthotharan Krishnan Nair

Reshmi Naduparambil Sasidharan

Thomas W Jacobs

Neeraj Jagdish Dhake

Ramanuja Narasimha Simha

Ghetia Siddharth

Jatin Varshney

Vikas Maddukuri

Laxmikant Mishra

Shaddy Garg

Yash Gadhia

Arjun Kashettiwar

Anjali Yadav

Shanka Subhra Mondal

Ravinder Nehra

In some embodiments, a computing system computes, with a state prediction model, probabilities of transitioning from a click state represented by interaction data to various predicted next states. The computing system computes an interface experience metric for the click with an experience valuation model. To do so, the computing system identifies base values for the click state and the predicted next states. The computing system computes value differentials for between the click state's base value and each predicted next state's base value. Value differentials indicate qualities of interface experience. The computing system determines the interface experience metric from a summation that includes the current click state's base value and the value differentials weighted with the predicted next states' probabilities. The computing system transmits the interface experience metric to an online platform, which can cause interface elements of the online platform to be modified based on the interface experience metric.

Machine-learning models applied to interaction data for facilitating experience-based modifications to interface elements in online environments

Some techniques described herein relate to utilizing a machine-learning (ML) model to select respective samples for queries of a query sequence. In one example, a method includes receiving a query in a query sequence, where the query is directed toward a dataset. Samples are available as down-sampled versions of the dataset. The method further include applying an agent to select, for the query, a sample from among the samples of the dataset. The agent includes an ML model trained, such as via intent-based reinforcement learning, to select respective samples for queries. The query is then executed against the sample to output a response.

Intent-aware learning for automated sample selection in interactive data exploration

The technology described herein is directed to a self-learning application scheduler for improved scheduling distribution of resource requests, e.g., job and service scheduling requests or tasks derived therefrom, initiated by applications on a shared compute infrastructure. More specifically, the self-learning application scheduler includes a reinforcement learning agent that iteratively learns a scheduling policy to improve scheduling distribution of the resource requests on the shared compute infrastructure. In some implementations, the reinforcement learning agent learns inherent characteristics and patterns of the resource requests initiated by the applications and orchestrates placement or scheduling of the resource requests on the shared compute infrastructure to minimize resource contention and thereby improve application performance for better overall user-experience.

Self-learning scheduler for application orchestration on shared compute cluster

A digital environment includes multiple computing nodes and a scheduling system that assigns workloads to computing nodes. The scheduling system includes an equivalence-class-based resource usage prediction system that receives a workload request and predicts an equivalence class for that workload request based on resource usage over time by the workload request or metadata associated with the workload request. The scheduling system also includes a workload assignment system that assigns the workload request to one or more of the computing nodes based on the predicted equivalence class. The number of equivalence classes is small relative to the total number of workloads that are scheduled (as an example, 10 to 15 equivalence classes for a total number of workloads in the tens or hundreds of thousands).

System and method for training and selecting equivalence class prediction modules for resource usage prediction

A method includes identifying interaction data associated with user interactions with a user interface of an interactive computing environment. The method also includes computing goal clusters of the interaction data based on sequences of the user interactions and performing inverse reinforcement learning on the goal clusters to return rewards and policies. Further, the method includes computing likelihood values of additional sequences of user interactions falling within the goal clusters based on the policies corresponding to each of the goal clusters and assigning the additional sequences to the goal clusters with greatest likelihood values. Furthermore, the method includes computing interface experience metrics of the additional sequences using the rewards and the policies corresponding to the goal clusters of the additional sequences and transmitting the interface experience metrics to the online platform. The interface experience metrics are usable for changing arrangements of interface elements to improve the interface experience metrics.

Machine-learning models applied to interaction data for determining interaction goals and facilitating experience-based modifications to interface elements in online environments

There is described a method and system in an interactive computing environment modified with user experience values based on behavior logs. An experience valuation system determines an experience value and an estimated experience value. The experience value is based on a current state of interaction data from a user session, based on a history of past events, and an estimation function defined by parameters to model the user experience values. The estimated experience value is determined based on, in addition to the current state and the estimation function, next states associated with the current state, and a reward function. The parameters of the estimation function are updated based on a comparison of the expected experience value and the estimated experience value. For another aspect, the method and system may further include a state prediction system to determine probabilities of transitioning that may be applied to determine the estimated experience value.

Characterizing and modifying user experience of computing environments based on behavior logs

In some embodiments, a model training system trains a sample generation model configured to generate synthetic data entries for a dataset. The sample generation model includes a prior model for generating an estimated latent vector from a partially observed data entry, a proposal model for generating a latent vector from a data entry of the dataset and a mask corresponding to the partially observed data entry, and a generative model for generating the synthetic data entries from the latent vector and the partially observed data entry. The model training system trains the sample generation model to optimize an objective function that includes a first term determined using the synthetic data entries and a second term determined using the estimated latent vector and the latent vector. The trained sample generation model can be executed on a client computing device to service queries using the generated synthetic data entries.

Query-oriented approximate query processing based on machine learning techniques

In some embodiments, interaction data associated with user interactions with a user interface of an interactive computing environment is identified, and goal clusters of the interaction data are computed based on sequences of the user interactions and performing inverse reinforcement learning on the goal clusters to return rewards and policies. Further, likelihood values of additional sequences of user interactions falling within the goal clusters are computed based on the policies corresponding to each of the goal clusters and assigning the additional sequences to the goal clusters with greatest likelihood values. Computing interface experience metrics of the additional sequences are computed using the rewards and the policies corresponding to the goal clusters of the additional sequences and transmitting the interface experience metrics to the online platform. The interface experience metrics are usable for changing arrangements of interface elements to improve the interface experience metrics.

Systems and methods provide for generating predictive models that are useful in predicting next-user-actions. User-specific navigation sequences are obtained, the navigation sequences representing temporally-related series of actions performed by users during navigation sessions. To each navigation sequence, a Recurrent Neural Network (RNN) is applied to encode the navigation sequences into user embeddings that reflect time-based, sequential navigation patterns for the user. Once a set of navigation sequences is encoded to a set of user embeddings, a variety of classifiers (prediction models) may be applied to the user embeddings to predict what a probable next-user-action may be and/or the likelihood that the next-user-action will be a desired target action.

Predictive analysis of target behaviors utilizing RNN-based user embeddings

Growing community of inventors

Chandigarh, India

Nikhil Sheoran