中科微点-全球专利检索

  1. Login
  2. Sign Up

Search Results

40 records (took 0.019 seconds)

    SMOOTH BLENDING OF MACHINE LEARNING MODEL VERSIONS
    1.
    发明申请
    SMOOTH BLENDING OF MACHINE LEARNING MODEL VERSIONS 有权

    公开(公告)号：US20250077959A1

    公开(公告)日：2025-03-06

    申请号：US18461703

    申请日：2023-09-06

    Applicant: AMPERITY, INC.

    Inventor： Yan YAN ,  Pranav Behari LAL ,  Nicholas RESNICK ,  Joyce GORDON

    IPC: G06N20/00

    Abstract: In some implementations, the techniques described herein relate to a method including: loading a current and a new model, the new model including the most recent version of the current model; computing a migration duration based on computed properties, namely the jitter in predictions between the current and the new models based on imputing the same inference data to both models; blending outputs of the current model with outputs of the new model according to weights computed for a current time step in the migration process; and serving new predictions using the new model when the migration duration expires.

    Predicting customer lifetime value with unified customer data
    3.
    发明授权
    Predicting customer lifetime value with unified customer data 有权

    公开(公告)号：US12198072B2

    公开(公告)日：2025-01-14

    申请号：US18390803

    申请日：2023-12-20

    Applicant: AMPERITY, INC.

    Inventor： Yan Yan ,  Aria Haghighi ,  Nicholas Resnick ,  Andrew Lim

    IPC: G06N5/04 ,  G06F16/23 ,  G06F16/24 ,  G06N20/00 ,  G06Q30/0201 ,  G06Q30/01 ,  G06Q30/0202

    Abstract: Disclosed are techniques for generating features to train a predictive model to predict a customer lifetime value or churn rate. In one embodiment, a method is disclosed comprising receiving a record that includes a plurality of fields and selecting a value associated with a selected field in the plurality of fields. The method then queries a lookup table comprising a mapping of values to aggregated statistics using the value and receives an aggregated statistic based on the querying. Next, the method generates a feature vector by annotating the record with the aggregated statistic. The method uses this feature vector as an input to a predictive model.

    Data structures for managing configuration versions of cloud-based applications
    4.
    发明授权
    Data structures for managing configuration versions of cloud-based applications 有权

    公开(公告)号：US11080043B1

    公开(公告)日：2021-08-03

    申请号：US16896844

    申请日：2020-06-09

    Applicant: Amperity, Inc.

    Inventor： Gregory Kyle Look

    IPC: G06F8/71 ,  G06F21/53

    Abstract: The present disclosure relates to methods and systems for applying version control of configurations to a software application, such as, a cloud-based application. Each version may be stored as a plurality of configuration nodes within a configuration tree structure. Version changes may lead to the creation or modification of configuration nodes. Configurations may be tested in a sandbox and undergo validation checks before being applied to the software application.

    PREDICTING CUSTOMER LIFETIME VALUE WITH UNIFIED CUSTOMER DATA
    5.
    发明申请
    PREDICTING CUSTOMER LIFETIME VALUE WITH UNIFIED CUSTOMER DATA 有权

    公开(公告)号：US20250148322A1

    公开(公告)日：2025-05-08

    申请号：US19017142

    申请日：2025-01-10

    Applicant: AMPERITY, INC.

    Inventor： Yan YAN ,  Aria HAGHIGHI ,  Nicholas RESNICK ,  Andrew LIM

    IPC: G06N5/04 ,  G06F16/23 ,  G06F16/24 ,  G06N20/00 ,  G06Q30/01 ,  G06Q30/0201 ,  G06Q30/0202

    Abstract: Disclosed are techniques for generating features to train a predictive model to predict a customer lifetime value or churn rate. In one embodiment, a method is disclosed comprising receiving a record that includes a plurality of fields and selecting a value associated with a selected field in the plurality of fields. The method then queries a lookup table comprising a mapping of values to aggregated statistics using the value and receives an aggregated statistic based on the querying. Next, the method generates a feature vector by annotating the record with the aggregated statistic. The method uses this feature vector as an input to a predictive model.

    QUERY RUNTIME FOR MULTI-LAYER COMPOSITION OF QUERIES
    6.
    发明申请
    QUERY RUNTIME FOR MULTI-LAYER COMPOSITION OF QUERIES 有权

    公开(公告)号：US20250086180A1

    公开(公告)日：2025-03-13

    申请号：US18462762

    申请日：2023-09-07

    Applicant: AMPERITY, INC.

    Inventor： Kailashnath Reddy KAVALAKUNTLA

    IPC: G06F16/2453

    Abstract: The present disclosure describes a system and method for optimizing SQL queries, specifically addressing challenges in handling and optimization of nested Common Table Expressions (CTEs). The system comprises a SQL optimization engine configured to receive SQL scripts from a SQL editor application and output optimized SQL to a query engine for execution on a database. The optimization engine utilizes three primary stages: a CTE normalization stage, a materialization stage, and a caching stage. The CTE normalization stage unnests nested CTEs into single-level CTEs. The materialization stage implements a materialized Create Table As Select (CTAS) strategy for materializing the base query. The caching stage enables reusability of the materialized base query across multiple queries, increasing efficiency and performance. This system provides technical solutions to enhance the capabilities of SQL engines that lack native support for nested CTEs, offering improved query performance and management of large datasets.

    Trimming blackhole clusters
    7.
    发明授权
    Trimming blackhole clusters 有权

    公开(公告)号：US12013855B2

    公开(公告)日：2024-06-18

    申请号：US18313753

    申请日：2023-05-08

    Applicant: AMPERITY, INC.

    Inventor： Yan Yan ,  Aria Haghighi ,  Joseph Christianson

    IPC: G06F16/2453 ,  G06F16/2457 ,  G06F16/28

    CPC classification number: G06F16/24542 ,  G06F16/24578 ,  G06F16/285

    Abstract: Disclosed are techniques for trimming large clusters of related records. In one embodiment, a method is disclosed comprising receiving a set of clusters, each cluster in the clusters including a plurality of records. The method extracts an oversized cluster in the set of clusters and performs a breadth-first search (BFS) on the oversized cluster to generate a list of visited records. The method terminates the BFS upon determining that the size of the list of visited records exceeds a maximum size and generates a new cluster from the list of visited records and adding the new cluster to the set of clusters. By recursively performing BFS traverse over the oversized cluster and extracting smaller new clusters from it, the oversized cluster is eventually partitioned into a set of sub-clusters with the size smaller than the predefined threshold.

    GENERATING AFFINITY GROUPS WITH MULTINOMIAL CLASSIFICATION AND BAYESIAN RANKING
    9.
    发明申请
    GENERATING AFFINITY GROUPS WITH MULTINOMIAL CLASSIFICATION AND BAYESIAN RANKING 有权

    公开(公告)号：US20230131884A1

    公开(公告)日：2023-04-27

    申请号：US17511946

    申请日：2021-10-27

    Applicant: AMPERITY, INC.

    Inventor： Andrew LIM ,  Joseph CHRISTIANSON ,  Joyce GORDON ,  Nicholas RESNICK ,  Yan YAN

    IPC: G06N7/00 ,  G06N20/20

    Abstract: The example embodiments are directed toward improvements in generating affinity groups. In an embodiment, a method is disclosed comprising generating probabilities of object interactions for a plurality of users, a given object recommendation ranking for a respective user comprising a ranked list of object attributes; calculating interaction probabilities for each user over a forecasting window; calculating affinity group rankings based on the probabilities of object interactions and the interaction probabilities for each user; and grouping the plurality of users based on the affinity group rankings.

Patent Agency Ranking