知搜-全球专利检索

  1. Login
  2. Sign Up

Search Results

57 records (took 0.026 seconds)

    Federated learning for multi-label classification model for oil pump management
    1.
    发明授权
    Federated learning for multi-label classification model for oil pump management 有权

    公开(公告)号：US12182771B2

    公开(公告)日：2024-12-31

    申请号：US17123088

    申请日：2020-12-15

    Applicant: INTERNATIONAL BUSINESS MACHINES CORPORATION

    Inventor： Nianjun Zhou ,  Dhavalkumar C. Patel ,  Anuradha Bhamidipaty

    IPC: G06Q10/20 ,  G06N3/098 ,  G06Q50/06

    Abstract: A computer implemented federated learning method of predicting failure of assets includes generating a local model at a local site for each of the cohorts and training the local model on local data for each of the cohorts for each failure type. The local model is shared with a central database. A global model is created based on an aggregation of a plurality of the local models from a plurality of the local sites. At each of the plurality of local sites, one of the global model and the local model is chosen for each of the cohorts. The chosen model operates on local data to predict failure of the assets. The utilized features include partitioning features of the assets into static features, semi-static features, and dynamic features, and forming cohorts of the assets based on the static features and the semi-static features.

    MULTI-LOCATIONAL FORECAST MODELING IN BOTH TEMPORAL AND SPATIAL DIMENSIONS
    2.
    发明申请
    MULTI-LOCATIONAL FORECAST MODELING IN BOTH TEMPORAL AND SPATIAL DIMENSIONS 有权

    公开(公告)号：US20230073564A1

    公开(公告)日：2023-03-09

    申请号：US17458728

    申请日：2021-08-27

    Applicant: International Business Machines Corporation

    Inventor： Zhengliang Xue ,  Bhavna Agrawal ,  Anuradha Bhamidipaty ,  Yingjie Li ,  Shuxin Lin

    IPC: G06F30/20

    Abstract: Temporal and spatially integrated forecast modeling includes generating a plurality of forecast models for a plurality of short-term to long-term time periods for a plurality of locations. Temporally integrating the plurality of forecast models sequentially over the plurality of time periods for the plurality of locations and spatially integrating the temporally integrated plurality of forecast models for each location hierarchically over the geographic areas. The forecast models are autoregressive distributed lag models with different explanatory variables for the short-term and long-term forecast models. The temporally integrating includes recursively integrating the plurality of forecast models over the time periods from the short-term to the long-term time periods and the spatially integrating includes recursively integrating the temporally integrated plurality of forecast models hierarchically from larger size geographic areas to smaller size geographic areas. The method includes optimizing the resultant spatially and temporally integrated forecast model based on a plurality of constraints.

    MODEL FIDELITY MONITORING AND REGENERATION FOR MANUFACTURING PROCESS DECISION SUPPORT
    3.
    发明申请
    MODEL FIDELITY MONITORING AND REGENERATION FOR MANUFACTURING PROCESS DECISION SUPPORT 有权

    公开(公告)号：US20220011760A1

    公开(公告)日：2022-01-13

    申请号：US16923148

    申请日：2020-07-08

    Applicant: International Business Machines Corporation

    Inventor： Nianjun Zhou ,  Dharmashankar Subramanian ,  Dhavalkumar C. Patel ,  Anuradha Bhamidipaty

    IPC: G05B23/02 ,  G05B13/04 ,  G05B13/02

    Abstract: Techniques for model fidelity monitoring and regeneration for manufacturing process decision support are described herein. Aspects of the invention include determining that an output of a regression model corresponding to a current time period of decision support for a manufacturing process is not within a predefined range of a historical process dataset, wherein the regression model was constructed based on the historical process dataset, and performing an accuracy and fidelity analysis on the regression model based on process data from the manufacturing process corresponding to a previous time period. Based on a result of the accuracy and fidelity analysis being below a threshold, a mismatch of the regression model as compared to the manufacturing process is determined. Based on determining the mismatch, a temporary regression model corresponding to the manufacturing process is generated, and decision support for the manufacturing process is performed based on the temporary regression model.

    Evaluating project maturity from data sources
    4.
    发明授权
    Evaluating project maturity from data sources 有权

    公开(公告)号：US10241786B2

    公开(公告)日：2019-03-26

    申请号：US15415943

    申请日：2017-01-26

    Applicant: International Business Machines Corporation

    Inventor： Anuradha Bhamidipaty ,  Evelyn Duesterwald ,  Andrew L. Frenkiel ,  Peter H. Westerink

    IPC: G06F9/44 ,  G06F8/77 ,  G06Q10/00

    Abstract: Techniques are provided for performing automated operations to determine maturity of a specified project. Information is received regarding each of a plurality of artifacts associated with the project, such as project documentation, source code repositories, and a tracked issue database for the project. A data sufficiency level associated with each provided artifact is determined, and each artifact is provided to one or more of multiple analysis engines. The analysis engines are executed to produce one or more weighted feature vectors for each of the artifacts associated with the specified project, and input to a prediction engine in order to provide a maturity rating for the project based on the weighted feature vectors.

    Artificial Intelligence Model Factory
    7.
    发明申请
    Artificial Intelligence Model Factory 有权

    公开(公告)号：US20250045624A1

    公开(公告)日：2025-02-06

    申请号：US18362726

    申请日：2023-07-31

    Applicant: International Business Machines Corporation

    Inventor： Dhavalkumar C. Patel ,  Vivek Sharma ,  Anuradha Bhamidipaty ,  Jayant R. Kalagnanam ,  Shuxin Lin ,  Dhruv Shah ,  Srideepika Jayaraman

    IPC: G06N20/00

    Abstract: An approach for generating an artificial intelligence system configurable for use with assets. In this approach, a model recipe is selected for generating the artificial intelligence system for use with assets. Recipe parameters specified in the model recipe are identified. A training dataset is created using the model recipe and input data. A set of artificial intelligence models is trained using the training dataset, the recipe parameters, and the model recipe. The training creates artifact models. The artifact models resulting from training are evaluated. The evaluation is used to select a set of the artifact models in the artifacts that form the artificial intelligence system that is configurable for use in assets.

    PIPELINE SELECTION FOR MACHINE LEARNING MODEL BUILDING
    8.
    发明申请
    PIPELINE SELECTION FOR MACHINE LEARNING MODEL BUILDING 有权

    公开(公告)号：US20240427604A1

    公开(公告)日：2024-12-26

    申请号：US18341120

    申请日：2023-06-26

    Applicant: International Business Machines Corporation

    Inventor： Chandrasekhara K. Reddy ,  Yuhan Shao ,  Dhavalkumar C. Patel ,  Jayant R. Kalagnanam ,  Anuradha Bhamidipaty

    IPC: G06F9/38

    Abstract: Machine learning (ML) pipeline selection includes performing cross-validation runs for dataset-pipeline combinations and building a matrix of first accuracy scores, factoring the matrix of accuracy scores into pipeline latent factors and dataset latent factors, augmenting the matrix of accuracy scores by selecting a subset of ML pipelines of a plurality of ML pipelines, then, for a new dataset, running the subset of ML pipelines with the new dataset to build and test respective ML models, obtain second accuracy scores, and augment the matrix of accuracy scores with the second accuracy scores to produce an augmented matrix of accuracy scores, factoring the augmented matrix of accuracy scores into refined pipeline latent factors and refined dataset latent factors, and identifying, based on the refined pipeline latent factors and the refined dataset latent factors, ML pipeline(s), of the plurality of ML pipelines, as most optimal for model building based on the new dataset.

Patent Agency Ranking