知搜-全球专利检索

  1. Login
  2. Sign Up

Search Results

49 records (took 0.02 seconds)

    COMPOUND MODEL FOR EVENT-BASED PROGNOSTICS
    11.
    发明公开
    COMPOUND MODEL FOR EVENT-BASED PROGNOSTICS 审中-公开

    公开(公告)号：US20230206111A1

    公开(公告)日：2023-06-29

    申请号：US17561397

    申请日：2021-12-23

    Applicant: Hitachi, Ltd.

    Inventor： Mahbubul ALAM ,  Dipanjan GHOSH ,  Ahmed FARAHAT ,  Laleh JALALI ,  Chetan GUPTA ,  Shuai Zheng

    IPC: G06N20/00

    CPC classification number: G06N20/00

    Abstract: Example implementations described herein can involve systems and methods involving, for receipt of input data from one or more assets, identifying and separating different event contexts from the input data; training a plurality of machine learning models for each of the different event contexts; selecting a best performing model from the plurality of machine learning models to form a compound model; selecting a best performing subset of the input data for the compound model based on maximizing a metric; and deploying the compound model for the selected subset.

    DEEP REINFORCEMENT LEARNING WITH SHORT-TERM ADJUSTMENTS
    12.
    发明申请
    DEEP REINFORCEMENT LEARNING WITH SHORT-TERM ADJUSTMENTS 有权

    公开(公告)号：US20220012585A1

    公开(公告)日：2022-01-13

    申请号：US16925538

    申请日：2020-07-10

    Applicant: Hitachi, Ltd.

    Inventor： Hamed KHORASGANI ,  Chi ZHANG ,  Susumu SERITA ,  Chetan GUPTA

    IPC: G06N3/08 ,  G06N3/04

    Abstract: Example implementations described herein involve a new reinforcement learning algorithm to address short-term goals. In the training step, the proposed solution learns the system dynamic model (short-term prediction) in a linear format in terms of actions. It also learns the expected rewards (long-term prediction) in a linear format in terms of actions. In the application step, the proposed solution uses the learned models plus simple optimization algorithms to find actions that satisfy both short-term goals and long-term goals. Through the example implementations, there is no need to design sensitive reward functions for achieving short-term and long-term goals concurrently. Further, there is better performance in achieving short-term and long-term goals compared to the traditional reward modification methods, and it is possible to modify the short-term goals without time-consuming retraining.

    MULTI TASK LEARNING WITH INCOMPLETE LABELS FOR PREDICTIVE MAINTENANCE
    14.
    发明申请
    MULTI TASK LEARNING WITH INCOMPLETE LABELS FOR PREDICTIVE MAINTENANCE 有权

    公开(公告)号：US20210048809A1

    公开(公告)日：2021-02-18

    申请号：US16540810

    申请日：2019-08-14

    Applicant: Hitachi, Ltd.

    Inventor： Chi ZHANG ,  Ahmed Khairy FARAHAT ,  Chetan GUPTA ,  Karan AGGARWAL

    IPC: G05B23/02 ,  G06N3/08

    Abstract: Example implementations described herein involve, for data having incomplete labeling to generate a plurality of predictive maintenance models, processing the data through a multi-task learning (MTL) architecture including generic layers and task specific layers for the plurality of predictive maintenance models configured to conduct tasks to determine outcomes for one or more components associated with the data, each task specific layer corresponding to one of the plurality of predictive maintenance models; the generic layers configured to provide, to the task specific layers, associated data to construct each of the plurality of predictive maintenance models; and executing the predictive maintenance models on subsequently recorded data.

    MANAGEMENT AND EXECUTION OF EQUIPMENT MAINTENANCE
    15.
    发明申请
    MANAGEMENT AND EXECUTION OF EQUIPMENT MAINTENANCE 审中-公开

    公开(公告)号：US20200075027A1

    公开(公告)日：2020-03-05

    申请号：US16121837

    申请日：2018-09-05

    Applicant: HITACHI, LTD.

    Inventor： Adriano Siqueira ARANTES ,  Marcos VIEIRA ,  Chetan GUPTA ,  Ahmed Khairy FARAHAT ,  Maria Teresa GONZALEZ DIAZ

    IPC: G10L17/22 ,  G06N99/00 ,  G10L13/04 ,  G10L17/00

    Abstract: In some examples, a system may receive from a device, speech sound patterns corresponding to a voice input related to equipment. Further, the system may determine an identity of a person associated with the device, and may identify the equipment related to the voice input. Using at least one of the received speech sound patterns or a text conversion of the speech sound patterns, along with an equipment history of the identified equipment, as input to one or more machine learning models, the system may determine, at least partially, an instruction related to the equipment. Additionally, the system may send, to the device, the instruction related to the equipment as an audio file for playback on the device.

    METHOD AND SYSTEM FOR LEAN ANGLE ESTIMATION OF MOTORCYCLES
    17.
    发明申请
    METHOD AND SYSTEM FOR LEAN ANGLE ESTIMATION OF MOTORCYCLES 审中-公开

    公开(公告)号：US20190271543A1

    公开(公告)日：2019-09-05

    申请号：US16332210

    申请日：2017-08-07

    Applicant: Hitachi, Ltd.

    Inventor： Susumu SERITA ,  Chetan GUPTA

    IPC: G01B21/22 ,  B62J99/00

    Abstract: Example implementations described herein are directed to a system for lean angle estimation without requiring specialized calibration. In example implementations, the mobile device sensor data can be utilized without any additional specialized data or configuration to estimate the lean angle of a motorcycle. The lean angle is determined based on a determination of a base attitude of a mobile device and a measured attitude of the mobile device.

    BOUND ENHANCED REINFORCEMENT LEARNING SYSTEM FOR DISTRIBUTION SUPPLY CHAIN MANAGEMENT
    18.
    发明公开
    BOUND ENHANCED REINFORCEMENT LEARNING SYSTEM FOR DISTRIBUTION SUPPLY CHAIN MANAGEMENT 审中-公开

    公开(公告)号：US20240265339A1

    公开(公告)日：2024-08-08

    申请号：US18105593

    申请日：2023-02-03

    Applicant: Hitachi, Ltd.

    Inventor： Haiyan WANG ,  Atsuki KIUCHI ,  Hsiu-Khuern TANG ,  Chetan GUPTA ,  Ibrahim EL-SHAR ,  Wenhuan SUN

    IPC: G06Q10/087 ,  G06Q30/0202

    CPC classification number: G06Q10/087 ,  G06Q30/0202

    Abstract: Example implementations described herein involve systems and methods for bound enhanced reinforcement learning systems for distribution supply chain management which can include initializing a replay buffer, a first state-action value function network having first random weights, and a second state-action value function network having second random weights; determining an action corresponding to an inventory ordering quantity at one or more facility in a multi-echelon supply chain network based on an (epsilon) ϵ-greedy exploration policy; executing the action in a simulated environment, and storing transition results in the replay buffer; calculating an upper bound and a lower bound of the optimal inventory costs; incorporating the upper bound and the lower bound with at least hyper-parameters T1, τ2 in updating at least one of the first or the second state-action value function networks; and performing a gradient descent on the first state-action value function network based on the upper or the lower bound.

    METHOD FOR TWO-WAY TIME SERIES DIMENSION REDUCTION
    19.
    发明公开
    METHOD FOR TWO-WAY TIME SERIES DIMENSION REDUCTION 审中-公开

    公开(公告)号：US20240185059A1

    公开(公告)日：2024-06-06

    申请号：US18075228

    申请日：2022-12-05

    Applicant: Hitachi, Ltd.

    Inventor： Aniruddha Rajendra RAO ,  Haiyan WANG ,  Chetan GUPTA

    IPC: G06N3/08 ,  G06N3/0455

    CPC classification number: G06N3/08 ,  G06N3/0455

    Abstract: Systems and methods described herein can involve training a functional encoder involving a plurality of layers of continuous neurons from input time series data to learn a dimension reduced form of the input time series data, the dimension reduced form of the input time series data being at least one of a feature reduced or time point reduced form of the input time series data; and training a functional decoder comprising another plurality of layers of continuous neurons to learn the input time series data from the dimension reduced form of the input time series data.

Patent Agency Ranking