知搜-全球专利检索

  1. 登录
  2. 注册

搜索结果

183 条记录 (耗时 0.033 秒)

    METHODS AND SYSTEMS FOR NEURAL ARCHITECTURE SEARCH
    2.
    发明公开
    METHODS AND SYSTEMS FOR NEURAL ARCHITECTURE SEARCH 审中-公开

    公开(公告)号：US20240354579A1

    公开(公告)日：2024-10-24

    申请号：US18732052

    申请日：2024-06-03

    申请人： Swisscom AG

    发明人： Yassine Benyahia ,  Kamil Bennani-Smires ,  Michael Baeriswyl ,  Claudiu Musat

    IPC分类号： G06N3/086 ,  G06F18/24 ,  G06F40/20 ,  G06N3/045 ,  G06N3/048

    CPC分类号： G06N3/086 ,  G06F18/24 ,  G06F40/20 ,  G06N3/045 ,  G06N3/048

    摘要： Methods and systems are provided for neural architecture search. In a system with suitable processing circuitry, a preferred model may be determined for performing a selected task, with the determining including obtaining a computational graph that includes a plurality of nodes and a corresponding plurality of weightings configured to scale input data into the nodes. The computational graph defines a first model and a second model with each of the models including a subgraph in the computational graph, with one or more of the plurality of weightings being shared between the first model and the second model. One or more weightings of each of the models may be updated based on training of each of the models to perform the selected task, and the preferred model may be identified based on an analysis of both models. A neural network for performing the selected task may be configured based on the preferred model.

    Search space exploration for deep learning
    5.
    发明授权
    Search space exploration for deep learning 有权

    公开(公告)号：US11989656B2

    公开(公告)日：2024-05-21

    申请号：US16935445

    申请日：2020-07-22

    申请人： International Business Machines Corporation

    发明人： Chao Xue ,  Yonggang Hu ,  Lin Dong ,  Ke Wei Sun

    IPC分类号： G06N3/086 ,  G06N3/045

    CPC分类号： G06N3/086 ,  G06N3/045

    摘要： Aspects of the invention include systems and methods to obtain meta features of a dataset for training in a deep learning application. A method includes selecting an initial search space that defines a type of deep learning architecture representation that specifies hyperparameters for two or more neural network architectures. The method also includes applying a search strategy to the initial search space. One of the two or more neural network architectures are selected based on a result of an evaluation according to the search strategy. A new search space is generated with new hyperparameters using an evolutionary algorithm and a mutation type that defines one or more changes in the hyperparameters specified by the initial search space, and, based on the mutation type, the new hyperparameters are applied to the one of the two or more neural networks or the search strategy is applied to the new search space.

    Training a student neural network to mimic a mentor neural network with inputs that maximize student-to-mentor disagreement
    10.
    发明授权
    Training a student neural network to mimic a mentor neural network with inputs that maximize student-to-mentor disagreement 有权

    公开(公告)号：US11961003B2

    公开(公告)日：2024-04-16

    申请号：US16923913

    申请日：2020-07-08

    申请人： Nano Dimension Technologies, Ltd.

    发明人： Eli David ,  Eri Rubin

    IPC分类号： G06N3/044 ,  G06N3/086

    CPC分类号： G06N3/086 ,  G06N3/044

    摘要： A device, system, and method is provided for training a new neural network to mimic a target neural network without access to the target neural network or its original training dataset. The target neural network and the new neural network may be probed with input data to generate corresponding target and new output data. Input data may be detected that generate a maximum or above threshold difference between the corresponding target and new output data. A divergent probe training dataset may be generated comprising the input data that generate the maximum or above threshold difference and the corresponding target output data. The new neural network may be trained using the divergent probe training dataset to generate the target output data. The new neural network may be iteratively trained using an updated divergent probe training dataset dynamically adjusted as the new neural network changes during training.