-
公开(公告)号：US20220076113A1
公开(公告)日：2022-03-10
申请号：US17016503
申请日：2020-09-10
Applicant: International Business Machines Corporation
Inventor： Li Cao , Ze Ming Zhao , Xiao Tian Xu , Yi Shan Jiang
Abstract: Embodiments of the present disclosure relate to weight matrix prediction. In an embodiment, a computer-implemented method is disclosed. The method comprises sending a candidate weight matrix of a neural network to one of a plurality of computing nodes comprised in a computing system to perform a testing iteration. The method further comprises receiving a testing loss value from the one of the plurality of computing nodes based on the testing iteration. The method further comprises evaluating whether the testing loss value is applicable. The method further comprises determining that the candidate weight matrix is available to be employed in a new formal iteration in response to the testing loss value being applicable. In other embodiments, a system and a computer program product are disclosed.
-
公开(公告)号：US20240403181A1
公开(公告)日：2024-12-05
申请号：US18204707
申请日：2023-06-01
Applicant: International Business Machines Corporation
Inventor： Fan Jing Meng , Hua Ye , Hong Xin Hou , Ze Ming Zhao , Xiao Tian Xu , Jin Chi He , Peng Li
Abstract: A computer-implemented method, a system and a computer program product for device failure detection are disclosed. In the method, phase-based predictions may be performed on a plurality of storage devices to determine a plurality of sampling scopes and corresponding sampling ratios. The respective sampling scopes may comprise at least one storage device of the plurality of storage devices. A sampling dataset may be obtained by selecting a group of storage devices from the respective sampling scopes with the corresponding sampling ratios. Device failure may be detected for the group of storage devices based on the sampling dataset.
-
公开(公告)号：US11748617B2
公开(公告)日：2023-09-05
申请号：US17016503
申请日：2020-09-10
Applicant: International Business Machines Corporation
Inventor： Li Cao , Ze Ming Zhao , Xiao Tian Xu , Yi Shan Jiang
CPC classification number: G06N3/08 , G06F18/217 , G06F18/25 , G06F18/285
Abstract: Embodiments of the present disclosure relate to weight matrix prediction. In an embodiment, a computer-implemented method is disclosed. The method comprises sending a candidate weight matrix of a neural network to one of a plurality of computing nodes comprised in a computing system to perform a testing iteration. The method further comprises receiving a testing loss value from the one of the plurality of computing nodes based on the testing iteration. The method further comprises evaluating whether the testing loss value is applicable. The method further comprises determining that the candidate weight matrix is available to be employed in a new formal iteration in response to the testing loss value being applicable. In other embodiments, a system and a computer program product are disclosed.
-
公开(公告)号：US20230031636A1
公开(公告)日：2023-02-02
申请号：US17387125
申请日：2021-07-28
Applicant: International Business Machines Corporation
Inventor： Lin Dong , Dong Xie , Jing Li , Guang Han Sui , Xiao Tian Xu
IPC: G06N5/04 , G06F11/34 , G06F16/901
Abstract: Aspects of the invention include systems and methods configured to provide simplified and efficient artificial intelligence (AI) model deployment. A non-limiting example computer-implemented method includes receiving an AI model deployment input having pre-process code, inference model code, and post-process code. The pre-process code is converted to a pre-process graph. The inference model and the post-process model are similarly converted to an inference graph and a post-process graph, respectively. A pipeline path is generated by connecting nodes in the pre-process graph, the inference graph, and the post-process graph. The pipeline path is deployed as a service for inference.
-
Patent Agency Ranking
公开(公告)号：US20190325276A1
公开(公告)日：2019-10-24
申请号：US15959749
申请日：2018-04-23
Applicant: International Business Machines Corporation
Inventor： Zhiwen Fu , Yu Song , Xiao Tian Xu
Abstract: A method for dividing, by a training system, a computational training work load of one or more neural network layers; pre-training the one or more neural network layers with a first class of image data sensitive to an original known dataset; generating a first weight file from the first layer of the neural network based on the first class of image data sensitive to the original known dataset; loading the one or more pre-trained neural network layers and the generated first weight file into at least one Internet of Things (IoT) device; stacking the one or more pre-trained neural network layers with the first layer of the neural network to form a new training system for an uploaded new dataset; adjusting the generated first weight file based on an input of one or more new classes of image data comprised in the uploaded new dataset to generate a new second weight file; inferencing an object class of new image data comprised on the uploaded new dataset using the generated new second weight file; and outputting the inferenced object class of the new image data.
-
公开(公告)号：US12204612B1
公开(公告)日：2025-01-21
申请号：US18343164
申请日：2023-06-28
Applicant: International Business Machines Corporation
Inventor： Ze Ming Zhao , Peng Hui Jiang , Xiao Tian Xu , Wenjing Liao , Zhi E. Zhang
IPC: G06F18/23213 , G06F16/28
Abstract: Embodiments of the present disclosure provide systems and methods for implementing self-bias detection based on performance and importance. A disclosed computer implemented method aggregates continuous input data through a K-means clustering algorithm to reduce the number of aggregated sub-group data pairs, enabling a reduced calculation time for computing bias and enhanced performance. The self-bias detection identifies a scale factor and a balance factor of aggregated sub-group data pairs, which indicate the importance of the detected bias.
-
公开(公告)号：US20240046097A1
公开(公告)日：2024-02-08
申请号：US17817662
申请日：2022-08-05
Applicant: International Business Machines Corporation
Inventor： De Gao Chu , Lin Dong , Xiao Tian Xu , Xue Yin Zhuang
CPC classification number: G06N3/082 , G06K9/6228 , G06K9/6262
Abstract: A computer-implemented method for compressing a machine learning model includes converting an input machine learning model into a standard machine learning model. The method further includes converting the standard machine learning model into a plurality of pruned machine learning models, each of the pruned machine learning models converted using a corresponding pruning ratio from a pruning ratio candidate list. The method further includes determining, for each of the pruned machine learning models, a size-to-error ratio. The method further includes selecting, based on the size-to-error ratio of the pruned machine learning models, a first pruning ratio from the pruning ratio candidate list. The method further includes generating a compressed machine learning model by compressing the input machine learning model using the first pruning ratio that is selected. The method further includes deploying the compressed machine learning model for production.
-
公开(公告)号：US20230036851A1
公开(公告)日：2023-02-02
申请号：US17386003
申请日：2021-07-27
Applicant: International Business Machines Corporation
Inventor： Li Cao , Ze Ming Zhao , Zhan Wei Wang , Xiao Tian Xu
Abstract: The present invention provides a computer-implemented method, a system, and a computer program product for path planning According to the computer-implemented method, a target discriminator is selected from a set of discriminators with different kernel sizes based on a target image obtained from an image capturing device. In this case, a confidence of the target image is determined using the target discriminator. The confidence indicates whether the target image contains a target object to be captured. Thereby, a movement indication for moving the image capturing device to capture the target object is determined based on the confidence of the target image.
-
公开(公告)号：US10902302B2
公开(公告)日：2021-01-26
申请号：US15959749
申请日：2018-04-23
Applicant: International Business Machines Corporation
Inventor： Zhiwen Fu , Yu Song , Xiao Tian Xu
Abstract: A method for dividing, by a training system, a computational training work load of one or more neural network layers; pre-training the one or more neural network layers with a first class of image data sensitive to an original known dataset; generating a first weight file from the first layer of the neural network based on the first class of image data sensitive to the original known dataset; loading the one or more pre-trained neural network layers and the generated first weight file into at least one Internet of Things (IoT) device; stacking the one or more pre-trained neural network layers with the first layer of the neural network to form a new training system for an uploaded new dataset; adjusting the generated first weight file based on an input of one or more new classes of image data comprised in the uploaded new dataset to generate a new second weight file; inferencing an object class of new image data comprised on the uploaded new dataset using the generated new second weight file; and outputting the inferenced object class of the new image data.
-
-
-
-
-
-
-
-
Search Results
9
records
(took 0.035 seconds)
