AUTOMATED MACHINE LEARNING MODEL FEEDBACK WITH DATA CAPTURE AND SYNTHETIC DATA GENERATION

    公开(公告)号:US20220335328A1

    公开(公告)日:2022-10-20

    申请号:US17235026

    申请日:2021-04-20

    Abstract: Systems and techniques that facilitate automated machine learning model feedback with data capture and synthetic data generation are provided. In various embodiments, a receiver component can receive electronic input identifying a deployed machine learning model. In various aspects, a listener component can retrieve from a data pipeline a data candidate that has been analyzed by the deployed machine learning model, an inference generated by the deployed machine learning model based on the data candidate, and an expert conclusion provided by a subject matter expert based on the data candidate. In various instances, a comparison component can compare the inference with the expert conclusion to determine whether the inference is consistent with the expert conclusion. In various cases, an augmentation component can, in response to a determination that the inference is not consistent with the expert conclusion, generate a set of synthetic training data based on the data candidate.

    ANNOTATION PIPELINE FOR MACHINE LEARNING ALGORITHM TRAINING AND OPTIMIZATION

    公开(公告)号:US20210035015A1

    公开(公告)日:2021-02-04

    申请号:US16528121

    申请日:2019-07-31

    Abstract: Techniques are provided for enhancing the efficiency and accuracy of annotating data samples for supervised machine learning algorithms using an advanced annotation pipeline. According to an embodiment, a method can comprise collecting, by a system comprising a processor, unannotated data samples for input to a machine learning model and storing the unannotated data samples in an annotation queue. The method further comprises determining, by the system, annotation priority levels for respective unannotated data samples of the unannotated data samples, selecting, by the system from amongst different annotation techniques, one or more of the different annotation techniques for annotating the respective unannotated data samples based the annotation priority levels associated with the respective unannotated data samples.

    Annotation pipeline for machine learning algorithm training and optimization

    公开(公告)号:US11475358B2

    公开(公告)日:2022-10-18

    申请号:US16527965

    申请日:2019-07-31

    Abstract: Techniques are provided for enhancing the efficiency and accuracy of annotating data samples for supervised machine learning algorithms using an advanced annotation pipeline. According to an embodiment, a method can comprise collecting, by a system comprising a processor, unannotated data samples for input to a machine learning model and storing the unannotated data samples in an annotation queue. The method further comprises determining, by the system, annotation priority levels for respective unannotated data samples of the unannotated data samples, selecting, by the system from amongst different annotation techniques, one or more of the different annotation techniques for annotating the respective unannotated data samples based the annotation priority levels associated with the respective unannotated data samples.

    ANNOTATION PIPELINE FOR MACHINE LEARNING ALGORITHM TRAINING AND OPTIMIZATION

    公开(公告)号:US20210034920A1

    公开(公告)日:2021-02-04

    申请号:US16527965

    申请日:2019-07-31

    Abstract: Techniques are provided for enhancing the efficiency and accuracy of annotating data samples for supervised machine learning algorithms using an advanced annotation pipeline. According to an embodiment, a method can comprise collecting, by a system comprising a processor, unannotated data samples for input to a machine learning model and storing the unannotated data samples in an annotation queue. The method further comprises determining, by the system, annotation priority levels for respective unannotated data samples of the unannotated data samples, selecting, by the system from amongst different annotation techniques, one or more of the different annotation techniques for annotating the respective unannotated data samples based the annotation priority levels associated with the respective unannotated data samples.

Patent Agency Ranking